Handbook of Hepato-Pancreato-Biliary Surgery PDF

Handbook of
Hepato-
Pancreato-Biliary
Surgery
0002164308.INDD 1 7/15/2014 12:39:34 PM

002164308.INDD 2 7/15/2014 12:39:34 PM
Handbook of
Hepato-
Pancreato-Biliary
Surgery
Editor
Nicholas J. Zyromski, MD
Associate Professor
Department of Surgery
Indiana University School of Medicine
Indianapolis, Indiana
0002164308.INDD 3 7/15/2014 12:39:34 PM

Acquisitions Editor: Keith Donnellan
Product Development Editor: Brendan Huffman
Production Project Manager: Joan Sinclair
Design Coordinator: Teresa Mallon
Senior Manufacturing Manager: Beth Welsh
Marketing Manager: Daniel Dressler
Prepress Vendor: SPi Global
Copyright © 2015 Wolters Kluwer
All rights reserved. This book is protected by copyright. No part of this book may be
reproduced or transmitted in any form or by any means, including as photocopies
or scanned-in or other electronic copies, or utilized by any information storage and
retrieval system without written permission from the copyright owner, except for
brief quotations embodied in critical articles and reviews. Materials appearing in
this book prepared by individuals as part of their official duties as U.S. government
employees are not covered by the above-mentioned copyright. To request permis-
sion, please contact Wolters Kluwer Health at Two Commerce Square, 2001 Market
Street, Philadelphia, PA 19103, via email at [email protected], or via our web-
site at lww.com (products and services).
9 8 7 6 5 4 3 2 1
Printed in China
Library of Congress Cataloging-in-Publication Data

Handbook of hepato-pancreato-biliary surgery / edited by Nicholas J. Zyromski.
p. ; cm.
Includes bibliographical references and index.
ISBN 978-1-4511-8501-0
I. Zyromski, Nicholas J., editor.
[DNLM: 1. Pancreatic Diseases–surgery–Handbooks. 2. Biliary Tract D iseases–
surgery–Handbooks. 3. Digestive System Surgical Procedures–Handbooks. 4. Liver
Diseases–surgery–Handbooks. WI 39]
RD546
617.5'562–dc23
2014005744
This work is provided “as is,” and the publisher disclaims any and all warranties,
express or implied, including any warranties as to accuracy, comprehensiveness, or
currency of the content of this work.
This work is no substitute for individual patient assessment based upon
healthcare professionals’ examination of each patient and consideration of, among
other things, age, weight, gender, current or prior medical conditions, medication
history, laboratory data and other factors unique to the patient. The publisher does
not provide medical advice or guidance and this work is merely a reference tool.
Healthcare professionals, and not the publisher, are solely responsible for the use
of this work including all medical judgments and for any resulting diagnosis and
treatments.
Given continuous, rapid advances in medical science and health information,
independent professional verification of medical diagnoses, indications, appropri-
ate pharmaceutical selections and dosages, and treatment options should be made
and healthcare professionals should consult a variety of sources. When prescribing
medication, healthcare professionals are advised to consult the product informa-
tion sheet (the manufacturer’s package insert) accompanying each drug to verify,
among other things, conditions of use, warnings and side effects and identify any
changes in dosage schedule or contradictions, particularly if the medication to be
administered is new, infrequently used or has a narrow therapeutic range. To the
maximum extent permitted under applicable law, no responsibility is assumed by
the publisher for any injury and/or damage to persons or property, as a matter of
products liability, negligence law or otherwise, or from any reference to or use by
any person of this work.
LWW.com
002164308.INDD 4 7/15/2014 12:39:34 PM

To my family: Jennifer, Anna, Jake, Maddy, J.R., and Sarah.
Thank you for defining my life so completely.
0002164308.INDD 5 7/15/2014 12:39:34 PM

CONTRIBUTORS
Thomas A. Aloia, MD
Assistant Professor
Department of Surgical Oncology
The University of Texas MD Anderson Cancer Center
Houston, Texas
Massimo Arcerito, MD
Fellow
Division of HPB and Advanced GI Surgery
University of Michigan
Ann Arbor, Michigan
Chandrakanth Are, MD
Associate Professor
Fellow Division of Surgical Oncology
Department of Surgery/Genetics, Cell Biology and Anatomy
University of Nebraska Medical Center
Program Director
General Surgery Residency
University of Nebraska Medical Center
Omaha, Nebraska
Marshall S. Baker, MD, MBA
Assistant Clinical Professor of Surgery
Division of Surgical Oncology
University of Chicago Pritzker School of Medicine
NorthShore University Health System
Chicago, Illinois
Chad G. Ball, MD, MSC, FRCSC, FACS
Assistant Professor
University of Calgary
Foothills Medical Center
Calgary, Alberta, Canada
vi
002164308.INDD 6 7/15/2014 12:39:34 PM

CONTRIBUTORS vii
Danielle A. Bischof, MD
Clinical Fellow
Johns Hopkins Hospital
Baltimore, Maryland
Jordan P. Bloom, MD
Resident
Massachusetts General Hospital
Boston, Massachusetts
Rebecca A. Busch, MD
Resident
University of Wisconsin School of Medicine and Public Health
Madison, Wisconsin
Eugene P. Ceppa, MD
Assistant Professor
Division of Hepatopancreatobiliary Surgery
Clifford S. Cho, MD
Assistant Professor
Section of Surgical Oncology
Division of General Surgery
University of Wisconsin School of Medicine and Public Health
Madison, Wisconsin
Kathleen K. Christians, MD
Professor of Surgery
Medical College of Wisconsin
Milwaukee, Wisconsin
Gregory A. Coté, MD, MS
Assistant Professor of Medicine
Division of Gastroenterology and Hepatology
Rachelle N. Damle, MD
Resident in General Surgery
University of Massachusetts Medical School
Worcester, Massachusetts
0002164308.INDD 7 7/15/2014 12:39:34 PM

viii CONTRIBUTORS
Jashodeep Datta, MD
University of Pennsylvania Perelman School of Medicine
Philadelphia, Pennsylvania
Mashaal Dhir, MBBS

Nebraska Medical Center
Omaha, Nebraska
Timothy R. Donahue, MD
Assistant Professor of Surgery
Department of Molecular and Medical Pharmacology
University of California Los Angeles
Los Angeles, California
Jeffrey A. Drebin, MD, PhD

John Rhea Barton Professor of Surgery and Chairman
Barish H. Edil, MD
Associate Professor of Surgery
Pancreas and Biliary Surgery Program Director
University of Colorado
Aurora, Colorado
Douglas B. Evans, MD
Professor and Chair
William E. Fisher, MD
Director of Elkins Pancreas Center
Professor
Michael E. DeBakey Department of Surgery
Baylor College of Medicine
Houston, Texas
Evan L. Fogel, MSc, MD, FRCP(C)

Professor of Clinical Medicine
Digestive and Liver Disorders
Indiana University Health
University Hospital
002164308.INDD 8 7/15/2014 12:39:34 PM

CONTRIBUTORS ix
Jonathan A. Fridell, MD, FACS

Division of Transplant Surgery
Benjamin N. Gayed, MD
Michael G. House, MD
Associate Professor
Matthew S. Johnson, MD
Professor of Radiology
Department of Radiology
Michael L. Kendrick, MD
Professor of Surgery and Chair
Division of Subspecialty Surgery
Mayo Clinic College of Medicine
Rochester, Minnesota
Robert L. King, MD
Fellow, Interventional Radiology
Wesley D. Leung, MD, MS

Fellow, Gastroenterology
Keith D. Lillemoe, MD
Surgeon-in-Chief
Massachusetts General Hospital
W. Gerald Austen Professor of Surgery
Harvard Medical School
Boston, Massachusetts
0002164308.INDD 9 7/15/2014 12:39:35 PM

x CONTRIBUTORS
Edward Lin, DO, MBA, FACS

Emory University School of Medicine
Atlanta, Georgia
Jason B. Liu, MD
General Surgery Resident
University of Chicago Pritzker School of Medicine
Chicago, Illinois
Mary A. Maluccio, MD
Associate Professor
Richard S. Mangus, MD, FACS

Assistant Professor of Surgery
Division of Transplant Surgery
Robert C.G. Martin II, MD, PhD

Sam and Lolita Weakley Endowed Chair in Surgical Oncology
Director of the Division of Surgical Oncology
Director of the Upper GI and HPB Multi-Disciplinary Clinical
Professor of Surgery and Academic Advisory Dean
University of Louisville
Louisville, Kentucky
Rebecca M. Minter, MD
Associate Professor of Medical Education
Associate Chair of Education
HPB and Advanced GI Surgery
University of Michigan
Ann Arbor, Michigan
Somala Mohammed, MD
Department of General Surgery
Baylor College of Medicine
Houston, Texas
002164308.INDD 10 7/15/2014 12:39:35 PM

CONTRIBUTORS xi
Katherine A. Morgan, MD, FACS

Associate Professor
Division of Gastrointestinal and Laparoscopic Surgery
Medical University of South Carolina
Charleston, South Carolina
David M. Nagorney, MD
Professor
Section of Hepatobiliary and Pancreatic Surgery
Division of Multispecialty General Surgery
Attila Nakeeb, MD, FACS

Chief section of HPB Surgery and Surgical Oncology
Alessandro Paniccia, MD
General Surgery Resident
Anschutz Medical Campus
University of Colorado Denver
Denver, Colorado
Timothy M. Pawlik, MD, MPH, PhD, FACS

Professor of Surgery and Oncology
Chief Division of Surgical Oncology
Johns Hopkins Hospital
Baltimore, Maryland
Prejesh Philips, MD
Surgical Oncology Fellow
University of Louisville
Louisville, Kentucky
Charles H.C. Pilgrim, MBBS (Hons), PhD

Fellow
0002164308.INDD 11 7/15/2014 12:39:35 PM

xii CONTRIBUTORS
Henry A. Pitt, MD
Chief Quality Officer
Temple University Health System
Associate Vice Dean for Clinical Affairs
Temple University School of Medicine
Howard A. Reber, MD
David Geffen School of Medicine at UCLA
Los Angeles, California
Kumaresan Sandrasegaran, MD
Associate Professor of Radiology
Indiana University School of Medicine and Clinical Sciences
Juan M. Sarmiento, MD, FACS

Emory University School of Medicine
Atlanta, Georgia
C. Max Schmidt, MD, PhD, MBA, FACS

Associate Professor of Surgery, Biochemistry, and Molecular Biology
Shimul A. Shah, MD, MHCM

Director of Liver Transplantation and Hepatobiliary Surgery
University of Cincinnati
Cincinnati, Ohio
Steven M. Strasberg, MD
Pruett Professor of Surgery
Washington University in St. Louis
St. Louis, Missouri
Temel Tirkes, MD
Assistant Professor of Radiology
002164308.INDD 12 7/15/2014 12:39:35 PM

CONTRIBUTORS xiii
Mark J. Truty, MD, MSc

Assistant Professor
Senior Associate Consultant
Section of Hepatobiliary and Pancreatic Surgery
Division of Multispecialty General Surgery
Ching-Wei D. Tzeng, MD
Hepatopancreatobiliary Surgery Fellow
University of Texas MD Anderson Cancer Center
Houston, Texas
Jean-Nicolas Vauthey, MD
Bessie McGoldrick Professor in Clinical Cancer Research
Chief of Liver Service
The University of Texas MD Anderson Cancer Center
Houston, Texas
Charles M. Vollmer Jr, MD

Joshua A. Waters, MD
Resident
Nicholas J. Zyromski, MD
Associate Professor
0002164308.INDD 13 7/15/2014 12:39:35 PM

FOREWORD
Do we really need another book on HPB surgery? After all, there are
the books by Blumgart and Clavien and all the other abdominal sur-
gery textbooks. Well, the answer for the focused readership of this
Handbook of Hepato-Pancreato-Biliary Surgery is yes—definitely!
While many of these other definite treatises certainly address the
multitude of problems, questions, controversies, and various intri-
cacies of the world of HPB, most all are long tomes, not immedi-
ately and readily readable, and, as always, being the true reference
sources, are more than 2 years behind, and are very daunting to
the reader looking for a concise, up-to-date “review” of relevant
anatomy (not the total Grant’s anatomic detail), acute and chronic
pancreatitis, the spectrum of pancreatic neoplasms (both benign
and malignant), portal hypertension, selected, surgically relevant
biliary and hepatic anatomy, benign and malignant hepatic sur-
gical disorders, and transplantation. Several other focused topics
include minimally invasive HPB procedures, relevant novel tech-
nology being used currently, and even endoscopic treatments of
HPB disorders.
What differentiates this working handbook are the following
aspects. First, the topics/chapters are carefully selected to address
relevant, real-world topics. Second, the chapters are relatively short,
concise and, most importantly, readable for the medical student, resi-
dent, fellow, and practicing surgeon. Note that the chapters are a bit
more than the classic bare-bones “handbook” and address a topic in
enough depth to provide a comprehensive but not exhaustive discus-
sion. My term for this book would be “comprehensive brevity.” Third,
the book focuses on the surgical practice of HPB, with an emphasis
on diagnosis, treatment, some technique, and complications. Fourth,
the reference list at the end of each chapter offers “selected” read-
ings rather than an exhaustive library. Fifth, the authors are generally
younger, enthusiastic, and more engaging than the more typical clas-
sic greybeards—but all are active clinical practitioners who under-
stand the problems.
So, do we need another book on HPB surgery? I say yes—it is
not a textbook but equally so more than a “pocket handbook” to
xiv
002164308.INDD 14 7/15/2014 12:39:35 PM

Foreword xv
put out fires that arise; best description is comprehensive brevity.

Congratulations to the editors for developing a truly useful, up-to-
date, focused, readable review of current HPB surgery.
Michael G. Sarr, MD
James C. Masson Professor of Surgery
Mayo Clinic
0002164308.INDD 15 7/15/2014 12:39:35 PM

PREFACE
Over the past decade, hepato-pancreato-biliary (HPB) surgery has

emerged as a mature specialty of general surgery. While a significant
percentage of HPB surgery problems are malignant, HPB surgeons
also treat a large number of benign disease conditions such as acute
and chronic pancreatitis, bile duct injury and benign bile duct stric-
ture, primary sclerosing cholangitis, benign liver lesions, cirrhosis,
and portal hypertension. Diagnosis, evaluation, and both operative
and nonoperative treatment of these patients are often undertaken
in a multidisciplinary format. Surgical decision making in these com-
plex patients is challenging and requires excellent judgment that
comes from experience. Treatment of many HPB conditions (such as
necrotizing pancreatitis, chemotherapy for pancreatic cancer) is an
actively evolving process.
Given this entire picture, the goal of this handbook is to pro-
vide an in-depth yet concise overview of contemporary treatment
of specific HPB surgical conditions. The target audience for the book
includes surgical residents in training, HPB surgery fellows, and gen-
eral surgeons in practice who may not encounter these complex HPB
surgical conditions in day-to-day practice. Others who will likely find
the text useful include parallel specialty practitioners like gastro-
enterologists, interventional radiologists, and medical oncologists
(i.e. those typically involved in multidisciplinary treatment of HPB
patients) who will gain benefit from understanding surgical perspec-
tive of these conditions.
The authors have been selected specifically for their expertise
in their chapter’s specific topics. As a whole, the author block rep-
resents substantial academic gravitas in HPB surgery. The depth of
their surgical experience is accessible through clear prose and will
be beneficial to all those invested in the care of HPB surgery patients.
xvi
002164308.INDD 16 7/15/2014 12:39:35 PM

TABLE OF CONTENTS
Contributorsvi
Forewordxiv
Prefacexvi
SECTION I: PANCREAS 1
1 Pancreatic Anatomy and Physiology 2

Somala Mohammed and William E. Fisher
2 Acute Pancreatitis 14
Benjamin N. Gayed and Nicholas J. Zyromski
3 Chronic Pancreatitis 27
Katherine A. Morgan
4 Pancreatic Adenocarcinoma 39
Jashodeep Datta, Jeffrey A. Drebin, and Charles M. Vollmer Jr
5 Pancreatic Cysts Including Intraductal Pancreatic

Mucinous Neoplasm (IPMN) 51
Joshua A. Waters and C. Max Schmidt
6 Pancreatic Neuroendocrine Neoplasms 62

Rebecca A. Busch and Clifford S. Cho
7 Technical Aspects of Pancreatic Surgery 76

Charles H. C. Pilgrim, Douglas B. Evans, and Kathleen K. Christians
8 Minimally Invasive Pancreatic Surgery 89

Michael L. Kendrick
9 Complications of Pancreatic Surgery 99

Timothy R. Donahue and Howard A. Reber
xvii
0002164308.INDD 17 7/15/2014 12:39:35 PM

xviii Table of Contents
SECTION II: LIVER 110
10 Liver Anatomy, Physiology, and Preoperative

Evaluation 111
Ching-Wei D. Tzeng and Jean-Nicolas Vauthey
11 Cirrhosis and Portal Hypertension 126

Rachelle N. Damle and Shimul A. Shah
12 Intraoperative Ultrasound of the Liver, Bile Ducts,

and Pancreas 140
Nicholas J. Zyromski
13 Energy Devices for Parenchymal Transection

in Liver Surgery 157
Prejesh Philips and Robert C.G. Martin II
14 Hepatocellular Carcinoma 169

Mashaal Dhir and Chandrakanth Are
15 Management of Patients with Colorectal Liver

Metastases 178
Danielle A. Bischof and Timothy M. Pawlik
16 Neuroendocrine and Noncolorectal Liver

Metastases 187
Mark J. Truty and David M. Nagorney
17 Benign Liver Tumors 204

Jason B. Liu and Marshall S. Baker
18 Complications of Liver Surgery Including

Postoperative Hepatic Insufficiency 221
Ching-Wei D. Tzeng and Thomas A. Aloia
19 Multidisciplinary Approach to Liver Oncology 233

Mary A. Maluccio
20 Laparoscopic Liver Surgery 243

Juan M. Sarmiento and Edward Lin
002164308.INDD 18 7/15/2014 12:39:35 PM

Table of Contents xix
SECTION III: BILIARY 253
21 Liver and Biliary Anatomy 254

Steven M. Strasberg
22 Infections in Biliary Surgery 271

Massimo Arcerito and Rebecca M. Minter
23 Gallbladder Disease 279

Henry A. Pitt
24 Sphincter of Oddi Dysfunction 300

Evan L. Fogel
25 Common Bile Duct Stones and Exploration 314

Attila Nakeeb
26 Choledochal Cyst 325

Alessandro Paniccia and Barish H. Edil
27 Biliary Malignancies 332

Michael G. House
28 Minimally Invasive Biliary Surgery 345

Eugene P. Ceppa
29 Bile Leak and Bile Duct Injury 354

Jordan P. Bloom and Keith D. Lillemoe
SECTION IV: MULTIDISCIPLINARY HPB CARE 372
30 Liver and Pancreas Transplantation 373

Richard S. Mangus and Jonathan A. Fridell
31 HPB Imaging (Including IR, Ultrasound, MRI,

CT, Nuclear Medicine) 387
Temel Tirkes and Kumaresan Sandrasegaran
32 Endoscopy in HPB Surgery Including ERCP and EUS 397

Wesley D. Leung and Gregory A. Coté
0002164308.INDD 19 7/15/2014 12:39:35 PM

xx Table of Contents
33 HPB Trauma 409

Chad G. Ball
34 Interventional Radiology Support of HPB Surgery 415

Robert L. King and Matthew S. Johnson
Index433
002164308.INDD 20 7/15/2014 12:39:35 PM

SECTION I
Pancreas
0002086371.INDD 1 7/12/2014 3:51:38 PM

1 Pancreatic Anatomy and
Physiology
Somala Mohammed and William E. Fisher
GROSS ANATOMY
Surgeons must thoroughly understand the anatomy of the pancreas and its
surrounding structures in order to minimize inadvertent injury and subse-
quent complications. In an adult, the pancreas weighs between 75 and 100 g
and is about 15 to 20 cm long. However, the shape, size, and texture of the
pancreas are quite variable. The pancreas is located in the retroperitoneum
and is obliquely oriented with the tail being more superior than the head.
The pancreas is covered with a fine connective tissue but lacks a true sero-
sal surface or capsule. Pain from inflammation of the pancreas is therefore
typically described as being located in the midepigastrium and penetrat-
ing to the back. The inflammatory process in acute pancreatitis can spread
inferiorly through the retroperitoneum down the perinephric and paracolic
gutters rather than diffusely throughout the peritoneal cavity.
The pancreas is commonly divided into five regions: the head, unci-
nate process, neck, body, and tail. The head is the thickest part of the pan-
creas and lies to the right of the superior mesenteric vessels. It is attached to
the second and third portions of the duodenum, and the two organs share a
common blood supply from the pancreaticoduodenal arcade. The anterior
surface of the pancreatic head is near the first portion of the duodenum
and the transverse mesocolon while the posterior surface is close to the
hilum and medial border of the right kidney, right ureter, the inferior vena
cava, right renal vein and artery, and the right crus of the diaphragm. In
the majority of patients, the common bile duct passes through the pancre-
atic head and is covered by varying amounts of parenchyma before joining
with the main duct of Wirsung and emptying into the second portion of the
duodenum.
The uncinate process is an extension of the pancreatic head. It passes
posterior, inferior, and slightly to the left from the head, staying adjacent to
the third and fourth portions of the duodenum. It then continues behind
the superior mesenteric vessels. On a sagittal section, the uncinate process
can be seen between the aorta and the superior mesenteric artery (SMA).
The uncinate process varies in size and shape. It may be absent in some, or
it may completely encircle the superior mesenteric vessels in others. Short
fragile vessels from the SMA and vein supply the uncinate process and must
be divided during resection of the pancreatic head. Pancreatitis or cancer
can make the attachments between the uncinate process and surrounding
mesenteric vessels difficult to separate intraoperatively.
The neck of the pancreas is defined as the portion of the gland located
anterior to the superior mesenteric vein (SMV) and splenoportal conflu-
ence. It is the thinnest portion of the gland and is covered anteriorly by
the pylorus. The second lumbar vertebra is just posterior to the neck of
the pancreas, which can be crushed against this bony structure with blunt
002086371.INDD 2 7/12/2014 3:51:38 PM

Chapter 1 / Pancreatic Anatomy and Physiology 3
a nterior–posterior abdominal trauma. The gastroduodenal artery passes

superiorly to inferiorly to the right of the pancreatic neck and at the infe-
rior margin of the pancreas gives rise to the right gastroepiploic artery and
then terminates as the superior pancreaticoduodenal artery. An ulcer in the
posterior duodenal bulb can erode through the posterior wall of the duode-
num into the gastroduodenal artery, which must sometimes be oversewn
to stop the bleeding. Behind the pancreatic neck at its inferior border, the
SMV joins the splenic vein and then continues toward the porta hepatis as
the portal vein (PV).
Once the gastrocolic omentum is divided, the body and tail of the
pancreas can be seen occupying the floor of the lesser sac, posterior to
the stomach. The celiac axis is located superior to the proximal body of
the pancreas with the hepatic artery coursing to the right, and the splenic
artery coursing to the left along the superior border of the body and tail
of the pancreas. The splenic artery is often tortuous and may pass directly
through the pancreatic parenchyma or even anterior to the pancreatic tail.
The transverse mesocolon attaches to the inferior edge of the body and tail
of the pancreas. The body of the pancreas begins to the left of the superior
mesenteric vessels. This portion of the pancreas measures approximately
4 to 5 cm in width and 1.5 to 2 cm in thickness. It lies anterior to the aorta
at the origin of the SMA and is anterior to the left renal vessels, the left crus
of the diaphragm, and the splenic vein. Resection of the pancreatic tail with
division at the neck is equivalent to a 60% to 70% resection, while division
at the proximal body is equivalent to a 50% to 60% resection.
The tail of the pancreas refers to the portion of the pancreas that is
anterior to the left kidney and left adrenal gland. In about 50% of patients,
the pancreatic tail extends into the hilum of the spleen and can thus be
injured during a splenectomy. The tail of the pancreas is attached to the
splenic flexure of the colon. Therefore, the pancreas can also be injured
during a left colectomy.
EMBRYOLOGY AND PANCREATIC DUCT ANATOMY

The pancreas is derived as an outpouching of the primitive foregut endo-
derm and is formed by the fusion of a ventral and dorsal pancreatic bud.
The ventral bud develops into the inferior portion of the head and the unci-
nate process while the dorsal bud develops into the body and tail. The ven-
tral bud rotates behind the duodenum from the right to the left and fuses
with the dorsal bud by the 6th to 8th week of gestation. The fusion of the
buds results in fusion of the two ductal systems in most individuals. The
embryonal ventral duct connects directly with the common bile duct and
becomes the main duct of Wirsung. The embryonal dorsal duct arises from
the duodenum and becomes the accessory duct of Santorini. With gut rota-
tion, the two ducts fuse in the head of the pancreas such that in most cases,
the majority of the pancreas drains through the duct of Wirsung into the
common channel formed from the bile duct and pancreatic duct. A “com-
mon channel” greater than 10 mm has been termed “pancreaticobiliary
maljunction” (PBM). Patients with PBM are felt to have an increased inci-
dence of biliary malignancy. The length of the common channel is variable.
The junction of the main duct and the accessory duct occurs at a major
bend in the main duct termed the “genu.”
Two pancreatic ducts that drain into the duodenum are the main duct
of Wirsung and the accessory duct of Santorini. Throughout the tail and
body, the main duct runs midway between the superior and inferior borders
of the pancreas, slightly closer to the posterior side. After passing into the
neck, the main duct travels inferior and posterior before joining the distal
0002086371.INDD 3 7/12/2014 3:51:38 PM

4 Section I / Pancreas
common bile duct and entering the second portion of the duodenum. The
pancreatic duct is usually only 2 to 4 mm in diameter, with its widest por-
tion in the head. Pancreatic cancer within the head of the pancreas typi-
cally obstructs both the bile duct and pancreatic duct resulting in jaundice,
and imaging shows dilation of both bile and pancreatic ducts, the “double
duct sign.” Pancreatic cancer in the body or tail of the gland often obstructs
the pancreatic duct only; patients typically present after months of vague
abdominal pain radiating to the back.
The normal pancreatic duct contains around 20 secondary branches,
which drain the tail, body, head, and uncinate process. These branches
enter the main duct at right angles and alternate in location on each side of
the duct. The pressure of the pancreatic ductal system is about twice that in
the common bile duct, and this pressure differential prevents reflux of bile
into the pancreatic duct. The muscle fibers around the ampulla form the
sphincter of Oddi, which controls the flow of pancreatic and biliary secre-
tions into the duodenum. The sphincter’s contraction and relaxation are
regulated by complex neural and hormonal factors including cholecysto-
kinin (CCK) from the duodenal mucosa, which causes sphincter relaxation.
The accessory duct of Santorini drains the superior and anterior
portions of the pancreatic head. In 60% of individuals, the accessory duct
enters separately into the duodenum via the minor papilla, which is located
approximately 2 cm proximal and anterior to the ampulla of Vater. In 30% of
individuals, either no minor papilla is present or there is a minor papilla but
the terminal portion of the accessory duct is diminutive in size and does
not permit the passage of pancreatic fluid. In 10% of patients, no connec-
tion is present between the accessory duct and the main duct. In the latter
group of patients, contrast medium injected into the main duct would not
delineate the anatomy of the minor duct. The normal pancreatic ductal sys-
tem is quite small. Two to three milliliters of contrast medium can fill the
main pancreatic duct and 7 to 10 mL can fill its branches and smaller ducts
as well. Injection of contrast material forcefully and at large volumes risks
distention of the ducts and postprocedural pancreatitis.
Various congenital anomalies can result from the failure of rotation or
fusion of the two pancreatic buds and their associated ducts. Pancreas divi-
sum is the most common congenital variant of the ductal system, occurring
in 5% to 15% of the population, and resulting from failure of the dorsal and
ventral buds to fuse. Subsequently, the duct of Santorini from the dorsal bud
drains most of the pancreas via the minor papilla. In these patients, the infe-
rior portion of the head and uncinate process continues to drain separately
through the duct of Wirsung via the major papilla (Fig. 1.1). Most patients
with pancreas divisum are asymptomatic. However, in some patients the
minor papilla may be inadequate to handle the flow of pancreatic fluid from
the majority of the gland. This theoretically leads to outflow obstruction and
potentially, pancreatitis.
Annular pancreas (AP) is an uncommon variant characterized by a
thin band of normal pancreatic tissue surrounding the second portion of
the duodenum. Annular pancreas occurs due to incomplete rotation of the
ventral pancreatic bud, so that it remains on the right side of the duodenum.
The incidence of AP is approximately 1 out of 20,000 individuals. More than
60% of patients with this anomaly present during the neonatal period with
features of gastric outlet obstruction. Many children with annular pancreas
have other congenital abnormalities such as Down syndrome and esopha-
geal or duodenal atresias. Children with AP typically present with duode-
nal obstruction (often diagnosed by prenatal ultrasound). This obstruction
is treated by duodenostomy, as dividing the pancreatic annulus may lead
to pancreatic fistula from dividing a main pancreatic duct. In contrast to
002086371.INDD 4 7/12/2014 3:51:38 PM

B
FIGURE 1.1 A. Normal pancreatic ductal anatomy. B. Failure of the ventral and dorsal
buds to fuse results in pancreas divisum, in which case the majority of the pancreas drains
through the duct of Santorini into the minor papilla. In these patients, the inferior portion of
the head and the uncinate process continue to drain separately through the duct of Wirsung
into the major papilla. (From Mulholland MW, Lillemoe KD, Doherty GM, et al. Greenfield’s
surgery: scientific principles and practice, 4th ed. Philadelphia, PA: Lippincott Williams &
Wilkins, 2005:1939.)
c hildren, adults presenting with AP often have different and more challeng-
ing pancreatobiliary disease such as pancreatitis or biliary obstruction.
ARTERIAL BLOOD SUPPLY

The pancreas has a rich blood supply derived from both the celiac trunk and
the SMA (Fig. 1.2). The celiac trunk gives rise to the splenic artery, the left
gastric artery, and the common hepatic artery. The common hepatic artery
gives rise to the gastroduodenal artery before continuing as the proper
hepatic artery. The gastroduodenal artery gives off the right gastric artery
0002086371.INDD 5 7/12/2014 3:51:41 PM

FIGURE 1.2 Arterial anatomy of the pancreas. (From Moore KL, Agur AM, Dalley AF.
Clinically oriented anatomy, 7th ed. Philadelphia, PA: Lippincott Williams & Wilkins,
2013:266.)
superior to the duodenum, travels inferiorly, anterior to the pancreatic neck

and posterior to the duodenum, and gives rise to the right gastroepiploic
artery at the inferior border of the duodenum, and then continues as the
superior pancreaticoduodenal artery. This branches into the anterior and
posterior superior pancreaticoduodenal arteries. The SMA gives rise to the
inferior pancreaticoduodenal artery, which also divides into anterior and
posterior branches. The pancreaticoduodenal arcades are always present
and form an extensive network of blood vessels that supply both the pancre-
atic head and the second and third portions of the duodenum. This shared
blood supply renders duodenal-preserving pancreatectomy a complex feat.
A rim of pancreatic tissue containing the arcade must be left intact.
Patients with celiac stenosis may derive all hepatic arterial blood
flow retrograde from the SMA via collateral pancreaticoduodenal arcades.
Therefore, it is prudent to temporarily occlude the gastroduodenal artery
prior to dividing this vessel during pancreatic head resection. This maneu-
ver ensures adequate antegrade hepatic arterial flow is present from the
celiac artery.
The neck, body, and tail of the pancreas are supplied by many branches
from the splenic artery and the SMA. The inferior pancreatic artery usually
arises from the SMA and runs to the left along the inferior border of the
body and tail of the pancreas, parallel to the splenic artery. Three vessels run
002086371.INDD 6 7/12/2014 3:51:43 PM

perpendicular to the long axis of the pancreatic body and tail and connect
the splenic artery and inferior pancreatic artery. They are, from medial to
lateral, the dorsal, great, and caudal pancreatic arteries. At the pancreatic
neck, the dorsal pancreatic artery arises from the splenic artery and gives
off both right and left branches. The right branch supplies the head of the
pancreas and usually joins the posterior arcade. The left branch passes
through the body and tail of the pancreas, often making connections with
branches of the splenic artery or left gastroepiploic artery. These arteries
form arcades within the body and tail of the pancreas, and account for the
rich blood supply of the organ. Detailed knowledge of the blood supply to
the neck, body, and tail of the pancreas is important when performing a dis-
tal pancreatectomy with splenic preservation or a central pancreatectomy.
Equally important is recognizing the broad arterial variability present.
Preoperative planning for patients with pancreatic cancer includes
high-quality computed tomography imaging to evaluate the primary
tumor or any sites of distant metastases, assess the patency of nearby ves-
sels, and delineate their relationship to the primary lesion. Tumors can
then be classified as resectable, locally advanced, or metastatic. A subset
of tumors blurs the distinction between resectable and locally advanced.
These tumors of borderline resectability include those that abut the SMA,
celiac axis, or hepatic artery (<180 degrees) or display short-segment occlu-
sion of the SMV, portal vein, or confluence of the two vessels with suitable
remaining vessels for reconstruction. Locally advanced, surgically unre-
sectable tumors include those that encase the celiac axis, hepatic artery, or
SMA (>180 degrees), or that occlude the SMV, portal vein, or its confluence
leaving no technical options for reconstruction. Encasement is defined as
involvement of greater than 50% of the circumference of the vessel whereas
abutment refers to less than 50% involvement.
Preoperative imaging also delineates aberrant vascular anatomy. The
most common arterial variant is a replaced right hepatic artery that arises
from the SMA instead of the proper hepatic artery. This variant is found in
10% to 15% of patients and usually courses posteriorly and superiorly from
the SMA around the posterior side of the portal vein and then up to the
porta hepatis on the right side. The artery may be involved by pancreatic
head tumors. It must also be differentiated from the inferior pancreatico-
duodenal arteries, which also arise from the SMA and take a similar course
to the replaced right hepatic. Inadvertent injury or resection of a replaced
right hepatic artery can lead to hepatic ischemia or compromise of the bili-
ary enteric anastomosis. A replaced left hepatic artery, present in 10% of the
population, typically arises from the left gastric artery and travels along the
superior border of the lesser omentum. A replaced left hepatic artery is usu-
ally distant from pancreatic head masses but may be involved with tumors
of the pancreatic body. Less common arterial variants include accessory
left and right hepatic arteries, which are similar to the replaced hepatic
arteries but are found in addition to the typical hepatic arterial anatomy.
VENOUS BLOOD SUPPLY

The venous drainage of the pancreas follows the arterial supply. The veins
are generally located anterior to the arteries, and both veins and arteries
lie posterior to the pancreatic ducts. All pancreatic veins ultimately drain
into the portal vein, splenic vein, SMV, or inferior mesenteric vein. Just as an
arterial arcade supplies the pancreatic head, a venous arcade of pancreati-
coduodenal vessels drains this region as well. The anterior superior pancre-
aticoduodenal vein joins the right gastroepiploic vein, which also receives
the middle colic vein, and drains directly into the SMV. Anterior traction on
0002086371.INDD 7 7/12/2014 3:51:43 PM

the transverse colon during colectomy or pancreatectomy can avulse these

veins, which then retract behind the pancreas and can be difficult to control.
The posterior superior pancreaticoduodenal vein enters the portal
vein above the superior margin of the pancreas. The anterior and posterior
inferior pancreaticoduodenal veins enter the SMVs together or separately.
Typically, numerous small venous branches drain from the pancreatic
parenchyma directly into the lateral posterior aspect of the portal vein.
With few exceptions, the veins generally enter along the lateral and poste-
rior sides of the SMV or portal vein and there are usually no anterior venous
tributaries. Therefore, a plane can be developed behind the pancreatic neck
and the underlying superior mesenteric/portal vein during pancreatic
resection, unless the tumor is invading the vein.
Three major venous branches drain the body and tail of the pancreas:
the inferior pancreatic vein, the caudal pancreatic vein, and the great pan-
creatic vein. All these vein branches drain into the splenic vein, which runs
in a groove on the dorsal pancreas. Many other unnamed venous branches
from the pancreatic parenchyma also drain into the splenic vein, and these
branches must be divided when performing a distal pancreatectomy with
splenic preservation. The inferior mesenteric vein courses behind the pan-
creas and usually joins the splenic vein. In some cases (about 15%), the
inferior mesenteric vein joins the left side of the SMV or directly with the
portal vein at the splenoportal confluence (Fig. 1.3). This variable anatomy
has implications for pancreaticoduodenectomy with venous resection.
Ideally the splenic vein–portal vein junction should be preserved if possible,
especially if the inferior mesenteric vein needs to be ligated and divided or
FIGURE 1.3 Venous anatomy of the pancreas. (From Moore KL, Agur AM, Dalley AF.
Clinically oriented anatomy, 7th ed. Philadelphia, PA: Lippincott Williams & Wilkins,
2013:266.)
002086371.INDD 8 7/12/2014 3:51:44 PM

if the inferior mesenteric vein enters the SMV. This may require reimplanta-
tion of the splenic vein into the interposition graft (usually the jugular vein).
Vascular involvement by tumor no longer represents an absolute con-
traindication to pancreatectomy. Cancer located at the inferior aspect of the
pancreatic head or at the uncinate process may involve either the portal vein
or the SMV with or without involvement of one of its two primary branches,
the jejunal and ileal veins. The infrapancreatic venous anatomy is variable. The
main trunk of the SMV is observed in over 90% of patients, but in the remain-
ing 10% of patients, the jejunal and ileal veins merge at the level of the splenic
vein without forming a common trunk. These veins are smaller and more
fragile than the main trunk, and either one of them can safely be ligated and
resected if the other is of sufficient caliber to allow for collateral venous drain-
age of the gut. Involvement of the confluence of the two smaller veins along
with the main trunk is managed by ligation of the jejunal branch as well as
segmental resection and reconstruction of the SMV trunk and proximal ileal
branch. Reconstruction of this branch is preferred because the jejunal branch
is usually more posterior and technically difficult to access for reconstruction.
LYMPHATIC DRAINAGE
A diffuse and widespread network of lymphatic channels is closely asso-
ciated with the blood vessels supplying the pancreas. Pancreatic cancer is
metastatic to the lymph nodes in 50% to 70% of patients. There are five main
groups of lymph nodes for the pancreas: superior, inferior, anterior, poste-
rior, and splenic nodes (Fig. 1.4). The anterior and posterior superior half of
the pancreatic head drains to the superior nodes, which are located along
the superior border of the pancreas and celiac trunk. The anterior and pos-
terior inferior half of the pancreatic head and body drain to inferior nodes,
located near the groove between the pancreas and duodenum. Additional
drainage is to nodes along the hepatoduodenal ligament, including those
along the portal vein and the hepatic artery.
The anterior portions of the pancreatic head also drain to a group
of anteriorly located nodes that ultimately drain to the right of the celiac
trunk and SMA, while the posterior portions of the head drain to posteriorly
located nodes. The upper part of the pancreatic body drains into the supe-
rior pancreatic nodes, while the lower part drains into inferior pancreatic,
superior mesenteric, and para-aortic nodes. Tumors of the body and tail
may be locally unresectable due to metastases to nodes in the transverse
mesocolon or jejunal mesentery. The tail of the pancreas drains to splenic
nodes located along the splenic vessels.
INNERVATION OF THE PANCREAS

Pancreatic innervation comes from the sympathetic division of the auto-
nomic nervous system through the splanchnic nerves and from the para-
sympathetic division through the vagus nerve. These nerves follow the
blood vessels and lymphatics. They contain a mixture of motor efferent and
sensory afferent nerve fibers from both autonomic systems. The parasym-
pathetic supply begins in the dorsal motor nucleus of the vagus nerve in
the medulla of the brain. These fibers travel in the vagus nerve and pass
through the celiac plexus. They then travel with arteries that branch from
the celiac trunk and ultimately enter the pancreatic parenchyma and syn-
apse with terminal ganglion cells within the gland. The postganglionic
fibers terminate at pancreatic islet cells. Almost 90% of the fibers carried by
the vagus nerve are sensory in function, related to stretch, chemoreceptors,
osmoreceptors, and thermoreceptors.
0002086371.INDD 9 7/12/2014 3:51:44 PM

FIGURE 1.4 Lymphatic drainage of the pancreas. (From Moore KL, Agur AM, Dalley
AF. Clinically oriented anatomy, 7th ed. Philadelphia, PA: Lippincott Williams & Wilkins,
2013:243.)
The sympathetic supply begins with ganglia in the thoracic spinal cord.
Fibers pass through the sympathetic chain and descend in the greater and
lesser splanchnic nerves. The former is composed of preganglionic efferent
fibers from the 5th through 9th thoracic segments while the latter is com-
posed of fibers from the 10th and 11th segments. These fibers pass through
the diaphragmatic crura and synapse on cell bodies in the celiac or superior
mesenteric ganglia. Postganglionic fibers then travel with branches of the
arteries to reach the pancreas.
Some afferent fibers cross midline in the celiac plexus. Interconnections
between afferent fibers from the pancreas and other sensory fibers from the
abdominal wall also exist. This anatomy may explain the referred pain associ-
ated with pancreatic diseases. The etiology of pain secondary to pancreatic
cancer is poorly understood, however. Explanations include infiltration of
nerve sheaths by malignancy, increased ductal pressure, and gland inflam-
mation. Pancreatic pain is generally transmitted through the celiac plexus,
located near the emergence of the celiac trunk from the aorta at the level of the
first lumbar vertebra. Celiac plexus nerve block can be performed for patients
with pain secondary to cancer as well as for patients with chronic pancreatitis.
PHYSIOLOGY
The pancreas is both an endocrine and an exocrine organ. The exo-
crine component of the pancreas accounts for 80% to 90% of the organ’s
mass, while the endocrine component accounts for approximately 2%.
002086371.INDD 10 7/12/2014 3:51:46 PM

The remainder of the pancreas consists of connective tissue, including the

extracellular matrix, blood vessels, and the ductal network. The exocrine
component secretes the enzymes responsible for digestion while the endo-
crine component is critical in glucose homeostasis.
The functional unit of the exocrine pancreas is the acinus and its
associated ductal system. Acinar cells are large pyramidal cells with their
basolateral aspect in contact with nerves, blood vessels, and the connective
tissue stroma and their apical aspect facing the central lumen of the aci-
nus. Within the apex are numerous zymogen granules that contain diges-
tive enzymes. Approximately 20 to 40 acinar cells come together to form
a functional unit called an acinus. The centroacinar cell, a second type of
cell in the acinus, secretes fluid and electrolytes to modify the pH of the
pancreatic fluid. The acinus drains into small intercalated ducts, which join
to form interlobular ducts and secondary ducts that ultimately drain into
the main pancreatic duct. Acute and chronic pancreatitis can cause duct
disruption which can lead to a pancreatic pseudocyst. Pancreatic inflam-
mation and subsequent scarring can also lead to duct stricture and dila-
tion, duct and acinar cell destruction, chronic pain, and eventually exocrine
pancreatic insufficiency.
The acinar cells secrete enzymes that fall into three major groups:
amylases, lipases, and proteases. Enzymatic secretion is stimulated by the
hormones secretin and CCK and by the parasympathetic nervous system.
Each acinar cell can secrete all types of enzymes, but the ratio of the differ-
ent enzymes released is adjusted to the composition of digested food. The
final product is an alkaline fluid that is colorless, odorless, and isosmotic,
containing over 20 enzymes and zymogens. Anywhere between 1 and 2 L
of fluid is secreted daily. The pH of this solution is approximately 8.0 and
results from the secretion of bicarbonate from centroacinar cells. Water
and electrolytes are also secreted by the centroacinar and intercalated cells
in response to secretin. Sodium and potassium cations are present in simi-
lar concentration as in plasma, but bicarbonate and chloride anions vary
in concentration according to their rate of secretion. As the rate increases,
bicarbonate concentration increases and chloride concentration decreases.
The sum of the two concentrations remains constant, however, and equals
that of plasma.
The endocrine function of the pancreas is performed by islets of
Langerhans, of which there are nearly 1 million in the normal adult pan-
creas. The islets consist of 3,000 to 4,000 cells of five major types: alpha cells
that secrete glucagon, beta cells that secrete insulin, delta cells that secrete
somatostatin, epsilon cells that secrete ghrelin, and PP or F cells that
secrete pancreatic polypeptide. The beta cells are located centrally within
the islet and constitute 70% of the islet mass whereas the other islet cell
types are located at the periphery. The PP, alpha, and delta cells account for
15%, 10%, and 5% of the islet cell mass, respectively. The cellular composi-
tion of the islets varies throughout the pancreas. Beta cells and delta cells
are present in all islets, whereas alpha cells are almost exclusively present
in the tail, body, and superior part of the head of the pancreas. PP cells are
almost exclusively present in the head of the pancreas (Table 1.1).
Although patients can live after total pancreatectomy with exogenous
digestive enzymes and hormones administration, the loss of islet–acinar
cell coordination leads to clinically challenging impairments in digestive
function and glucose regulation. Frequent glucose measurement and insu-
lin administration is required to correct for complete absence of insulin
production. However, these patients also lack the insulin counterregulatory
hormones and often have problems with severe hypoglycemia after insulin
administration. Only approximately 20% of the normal pancreas is required
0002086371.INDD 11 7/12/2014 3:51:46 PM

002086371.INDD 12
TABLE
Cells of the endocrine pancreas
1.1
Location Within Islet % of Islet Mass Location on Pancreas Hormone Secreted Hormone Function
Alpha Peripheral 10 Tail, body, superior head Glucagon ↑ Glycogenolysis
↑ Gluconeogenesis
Beta Central 70 Throughout Insulin ↑ Glycogen synthesis
↑ Protein synthesis
↑ Lipogenesis
Delta Peripheral 5 Throughout Somatostatin ↓ Endocrine/exocrine pancreas secretions
↓ Bile flow, gallbladder contraction
↓ Absorption of glucose, fats, amino acids
Many other functions
Epsilon Peripheral < 1 Throughout Ghrelin ↑ Hunger, growth hormone secretion
PP/F Peripheral 15 Head Pancreatic polypeptide ↑ Satiety
↑ Insulin release
The islets of Langerhans are composed of cells of five major types, depicted above.
7/12/2014 3:51:46 PM
to prevent exocrine or endocrine insufficiency after partial pancreatec-

tomy; however, many patients requiring pancreatectomy have diseased
pancreas remnant, and endocrine or exocrine insufficiency can develop
with removal of even smaller portions of the pancreas.
Suggested Readings
Balachandran A, Darden DL, Tamm EP, et al. Arterial variants in pancreatic adenocar-
cinoma. Abdom Imaging 2008;33(2):214–221.
Bockman DE. Anatomy and fine structure. In Beger HG, Warshaw AL, Buchler MW,
et al. (eds.), The pancreas, 2nd ed. Massachusetts, United States: Blackwell
Publishing, 2008.
Fisher WE, Andersen DK, Bell RH, et al. Pancreas. In Brunicardi FC, Andersen DK,
Billiar TR, et al. (eds.), Schwartz’s principles of surgery, 9th ed. New York, NY: The
McGraw-Hill Companies, 2010.
Glasgow RE, Mulvihill SJ. Liver, biliary tract, and pancreas. In: O’Leary JP (ed.),
Physiologic basis of surgery, 4th ed. Philadelphia, PA: Lippincott Williams &
Wilkins, 2008.
Katz MHG, Fleming JB, Pisters PWT, et al. Anatomy of the superior mesenteric vein
with special reference to the surgical management of first-order branch involve-
ment at pancreaticoduodenectomy. Annals Surg 2008;248(6):1098–1102.
Kooby DA, Loukas M, Skandalakis LJ, et al. Surgical anatomy of the pancreas. In:
Fischer JE (ed.), Fischer’s mastery of surgery, 6th ed. Philadelphia, PA: Lippincott
Williams & Wilkins, 2012.
Riall TS. Pancreas anatomy and physiology. In: Mulholland MW, Lillemoe KD, Doherty
GM, et al. (eds.), Greenfield’s surgery: scientific principles and practice, 5th ed.
Philadelphia, PA: Lippincott Williams & Wilkins, 2010.
Skandalakis JE, Skandalakis LJ, Kingsnorth AK, et al. Pancreas. In: Skandalakis JE,
Colborn GL, Weidman TA, et al. (eds.), Skandalakis’ surgical anatomy, 1st ed.
Athens, Greece: Paschalidis Medical Publications, 2004.
Varadhachary GR, Tamm EP, Abbruzzese JL, et al. Borderline resectable pancreatic
cancer: definitions, management, and role of preoperative therapy. Annals Surg
Oncol 2006;13(8):1035–1046.
0002086371.INDD 13 7/12/2014 3:51:47 PM

2
Acute Pancreatitis
Benjamin N. Gayed and Nicholas J.
Zyromski
INTRODUCTION
The incidence of acute pancreatitis (AP) in the United States has increased
over several decades, and current estimates exceed 40/100,000. This serious
medical condition accounts for more than 270,000 inpatient admissions in
the United States each year. Approximately 80% of AP cases in the United
States are mild and self-limited; the remaining 20% of pancreatitis cases
qualify as severe acute pancreatitis (SAP). Persistent organ failure is a defin-
ing feature of SAP, which is associated with high morbidity and a mortality
rate of 15% to 20%. Patients with SAP have variable necrosis of the pancre-
atic parenchyma and peripancreatic soft tissue. The natural history of SAP
is dynamic. The early phase lasts for the 1st week of AP; the late phase over-
laps with the early phase and lasts weeks to months. Up to 20% of people
with one episode of AP will go on to have chronic pancreatitis. This chapter
discusses the diagnosis and management of AP for surgeons: the majority
of the chapter is dedicated to the management of SAP and necrotizing pan-
creatitis (NP).
PATHOPHYSIOLOGY AND ETIOLOGY

The pathophysiology of pancreatitis is incompletely understood. Increased
resistance to pancreatic duct outflow from obstruction (as in gallstone
pancreatitis) or decreased radius (as in pancreas divisum) is one of several
pathophysiologic theories. Proposed theories all converge with premature
enzyme activation in the acinar cells and uninhibited activity of proteases
in the parenchyma. Etiologic factors in pancreatitis are better defined,
though regional and demographic variation exists. Table 2.1 summarizes
common etiologies of AP in the United States.
DIAGNOSIS AND ESTABLISHING SEVERITY

The diagnosis of AP is made clinically and is a diagnosis of exclusion.
Historical, biochemical, and radiographic factors support the diagnosis;
several of these factors help predict disease severity early in the disease
course. The classic presentation of AP involves acute onset of epigastric
pain, the intensity of which is usually severe and may radiate either directly
to the back or around the patient’s side. Patients often describe a “stabbing”
“knife-like” pain. Pain from AP is expected to be more centrally located
than is the pain of biliary colic, though the length of the pancreas, its posi-
tion across the retroperitoneum, and its visceral innervation may produce
symptoms that localize predominantly on either side of the abdomen. The
diagnosis of AP can be made when two of the following three features are
present:
14
002164307.INDD 14 7/14/2014 5:03:57 PM

Chapter 2 / Acute Pancreatitis 15
TABLE
2.1 Etiology of Acute Pancreatitis
Biliary
Ethanol abuse
Idiopathic
Iatrogenic (post-ERCP most common)
Medication (azathioprine, anti-HIV agents, and others)
Trauma
Hypercalcemia
Congenital
• Pancreas divisum
• Annular pancreas
Toxins
• Jamaican scorpion venom
Hereditary
Hypertriglyceridemia
Tumors
• IPMN (associated with a 25%–30% risk)
• Ductal adenocarcinoma
Other (autoimmune, ischemia, vasculitis)
1. Abdominal pain consistent with the diagnosis

2. Serum amylase or lipase at least three times the upper limit of normal
3. Characteristic imaging findings on contrast-enhanced computed tomo
graphy (CECT), magnetic resonance imaging (MRI), or transabdominal
ultrasound (U/S)
The onset of AP should be measured from the time that symptoms begin
rather than the time of clinical presentation. Elevated serum amylase and
lipase are not pathognomonic for AP and may be seen with other gastro-
intestinal (GI) conditions such as peptic ulcer disease, bowel obstruction,
perforation, etc. In chronic pancreatitis patients, these enzymes are less
sensitive diagnostic markers and may even be within normal range during
an acute flare. Once a diagnosis of AP has been established, it is appropriate
to consider the severity of disease.
Severity of AP
Identifying the severity of AP informs the need for closer monitoring, more
aggressive resuscitation, and identifying patients appropriate for transfer
to tertiary care centers. The evolution of AP severity must be anticipated
after diagnosis of AP. Careful and repeated examination is important early
in the disease course to determine clinical trajectory.
Multiple systems have been used to predict AP severity. Examples
of these scoring systems include Ranson’s criteria (Table 2.2), modi-
fied Glasgow score, Acute Physiologic and Chronic Health Evaluation
(APACHE II) score, the Sequential Organ Failure Assessment (SOFA)
score, the Marshall score, and the Balthazar score. No one scoring system
is universally accepted and applied in clinical practice. Recent evidence-
based guidelines suggest that measurement of the systemic inflammatory
response syndrome (SIRS) at admission and at 48 hours may be the easiest
and most practical marker of AP severity.
Various biochemical compounds have been studied as potential
markers of AP severity (Table 2.3). Serum C-reactive protein (CRP) is the
0002164307.INDD 15 7/14/2014 5:03:57 PM

TABLE
2.2
On Admission
Ranson’s Criteria
Within 48 h
Age > 55 y Calcium < 8.0 mg/dL
WBC > 16,000 cells/mm3 Hematocrit fall of ≥10%
Blood glucose > 200 mg/dL PO2 < 60 mm Hg
Serum AST > 250 IU/L BUN increase ≥5 mg/dL after fluid resuscitation
Serum LDH > 350 IU/L Base deficit >4 mEq/L
Fluid sequestration >6 L
Total score (30-d mortality):
0–2 (2%)
3–4 (15%)
5–6 (40%)
7–8 (100%)
most widely used; levels greater than 150 predict severe AP. Serum amy-
lase is not predictive of disease severity regardless of degree of elevation or
duration of elevation. Circulating amylase above 1,000 mg/dL supports the
diagnosis of a biliary etiology. Urinary trypsin activating peptide (TAP) and
interleukins (IL)-2 and 6 increase within 24 hours of AP onset, but clinical
utility is limited by the fact that these are typically send-out labs.
The presence and persistence of organ failure may be the single most
important predictor of outcome in SAP patients.
The Atlanta classification (developed in 1992 and revised in 2012)
identifies three grades of severity based on the presence and timing of
complications: mild, moderately severe, and severe. The revised Atlanta
criteria are summarized in Table 2.4. A new international “determinant-
based” classification has been proposed based on local and systemic
determinants of severity—that is, presence of (peri)pancreatic necrosis,
presence of infection in necrosis, and presence and persistence of organ
failure.
Imaging
Imaging in AP is used (1) to support the diagnosis, (2) to identify the pres-
ence of suspected complications, (3) to monitor disease progression (i.e.,
peripancreatic fluid collections), and (4) for operative planning.
Ultrasound
Transabdominal ultrasound is most useful to identify gallstones with
AP. It can be used in place of computed tomography (CT) to support the
diagnosis in the setting of normal amylase/lipase levels or with atypical
TABLE
2.3
Biochemical Markers of Acute
Pancreatitis Severity
Serum CRP
Interleukin-2
Interleukin-6
Procalcitonin
Polymorphonuclear elastase
Urinary trypsin activating peptide
002164307.INDD 16 7/14/2014 5:03:58 PM

TABLE
2.4 Revised Atlanta Criteria (2012)
Disease Severity Clinical Features

Mild acute pancreatitis • No organ failure
• No local or systemic complications
Moderately severe acute • Transient organ failure (resolves within 48 h)
pancreatitis • Local or systemic complications without
persistent organ failure
Severe acute pancreatitis • Persistent organ failure (>48 h) of one or more
organs
symptoms. Endoscopic ultrasound (EUS) can differentiate solid and cystic

components of retroperitoneal collections. Differentiating a true pseu-
docyst from walled-off pancreatic necrosis (WOPN) determines whether
a drainage procedure or debridement is indicated in a symptomatic or
“smoldering” patient (Fig. 2.1). EUS can also be used in cases of idiopathic
pancreatitis to better evaluate pancreatic parenchyma and ductal anatomy.
Intraoperative ultrasound is very helpful localizing peripancreatic collec-
tions, especially those that are predominantly solid.
Computed Tomography
CT is widely available and quite reproducible; as such, CT is the imaging
modality of choice in AP. Ideally, CT should be delayed 48 hours to provide
time for adequate resuscitation (to minimize nephrotoxicity) and because
early in the disease course, radiologic changes of AP are minimal. CECT
identifies the presence and extent of (peri)pancreatic fluid collections and
necrosis, hemorrhage and pseudoaneurysms (PSAs) (when timed for vis-
ceral arteriography), signs of infected necrosis (i.e., air in the retroperito-
neal collection), and venous thrombosis. CT is useful for monitoring the
disease progression as well, particularly for monitoring the size of fluid col-
lections, pseudocysts, and the extent of necrosis.
No prescribed timing exists to obtain follow-up CT imaging through
the course of severe AP, though many experienced practitioners fol-
low serial cross-sectional images on roughly a weekly basis until the
FIGURE 2.1 Ultrasound demonstrating WOPN—collection containing both

fluid and solid (arrows) necrosis.
0002164307.INDD 17 7/14/2014 5:03:58 PM

patient’s clinical condition stabilizes. Acute changes in clinical condi-

tion consistent with sepsis should prompt the provider to consider CT
imaging to evaluate for signs of infected pancreatic necrosis. Similarly,
clinical changes suggesting hemorrhage should prompt consideration of
arteriography.
Magnetic Resonance Imaging

MRI can be used in place of CT for diagnosis or surveillance. This modality
is particularly useful for patients with allergies to iodinated radiocontrast
or concerns about cumulative radiation dose. MRI is also perhaps the most
accurate cross-sectional imaging technique with which to distinguish solid
and cystic components of retroperitoneal fluid collections (Fig. 2.2). MRI
may also be combined with cholangiopancreatography (MRCP) to evalu-
ate the presence of bile duct stones and pancreatic duct integrity. The pres-
ence of parenchymal necrosis or a peripancreatic fluid collection obscures
MRCP evaluation of pancreatic duct architecture.
MANAGEMENT
Mild AP
Aggressive fluid resuscitation is the mainstay of therapy for mild AP. The
degree of resuscitation needed secondary to significant retroperitoneal
third spacing is often underappreciated; resuscitation should be guided
by evidence of end-organ perfusion (i.e., urine output). Pain is frequently
severe enough to warrant narcotic analgesics. Prophylactic antibiotic
administration is not indicated mild AP. Nasogastric (NG) decompression
may be used to relieve nausea and vomiting, but routine NG decompression
does not alter disease course. Gastric antisecretory medication (i.e., hista-
mine receptor-2 blockers, proton pump inhibitors) should be administered
routinely. Reintroduction of oral diet is appropriate when pain diminishes.
Up to 15% of patients may experience “refeeding pancreatitis” after oral
feeding is started; withholding oral diet for a short time is generally suc-
cessful therapy in these patients.
FIGURE 2.2 Magnetic resonance image of the same patient as Figure 2.1;
fluid and solid (arrow) necrosis.
002164307.INDD 18 7/14/2014 5:03:59 PM

Cholecystectomy is indicated in all cases of biliary pancreatitis and

should be considered in patients with idiopathic AP, or those with recurrent
pancreatitis initially attributed to another etiology. The biliary tree should
be interrogated by ERCP, MRCP, or intraoperative cholangiography in ALL
patients with AP. Removing the gallbladder more than 6 weeks after an initial
case of mild biliary pancreatitis leads to recurrent pancreatitis in over one-
third of patients. Because of this risk, cholecystectomy should be completed
at the earliest opportunity after pancreatitis resolution, ideally prior to hos-
pital discharge. Alternate arrangements for interval cholecystectomy shortly
following hospital discharge may be appropriate for responsible patients on a
case-by-case basis. In patients with NP, cholecystectomy may be deferred until
treatment of the necrotic collections becomes apparent. Should debridement
be necessary, cholecystectomy may be performed at the same setting.
Recurrence of AP is very low (3%) following endoscopic sphincterot-
omy (ES); however, other biliary symptoms (common bile duct stones, cho-
lecystitis, cholangitis) will occur in 10% to 15% of these patients. Therefore,
ES may be “definitive” treatment to prevent recurrent AP in patients who
are too infirm to tolerate cholecystectomy.
Severe Acute Pancreatitis and Necrotizing Pancreatitis

Natural History
The presence and persistence of organ failure and (peri)pancreatic necrosis
define SAP and NP. The mortality of patients with SAP/NP is approximately
20%—this is a serious medical problem. Patients with SAP/NP often require
long (>4- to 6-week) hospitalization, as well as some sort of intervention
to treat infected necrosis or local complications. Patients and families
should be counseled accordingly to expect long hospitalizations, “bumps
in the road” and perhaps up to several months of recuperation. Once necro-
sis becomes established, one of three outcomes is possible (Fig. 2.3). In a
small number of patients, the necrosis will reabsorb with no further conse-
quence. In a second group, the necrosis becomes infected, which typically
demands treatment. The third group of patients has persistent (presump-
tively sterile) necrosis. If asymptomatic, these patients do not need further
(PERI)PANCREATIC
COLLECTION
4 weeks
RESOLUTION
PERSISTS
INFECTION
YES SX?
INTERVENTION NO
OBSERVE
FIGURE 2.3 Outcomes after development of pancreatic necrosis.
0002164307.INDD 19 7/14/2014 5:04:00 PM

treatment directed to the necrotic collection. Symptomatic patients, on

the other hand, typically require intervention (these are the “persistently
unwell” patients who have nausea, epigastric or upper abdominal pain, low-
grade fevers, weight loss, etc.). It is noteworthy that up to 42% of patients
with presumed sterile necrosis will be found to have infected necrosis at the
time of pancreatic debridement.
The natural history of SAP occurs in two phases: early and late. Each
phase demonstrates a mortality peak. The early phase ends 1 to 2 weeks
after disease onset and is characterized by the release of proinflamma-
tory mediators and the resultant SIRS. The proinflammatory cascade and
systemic release of proteases both likely contribute to distant organ fail-
ure; renal and pulmonary systems are particularly vulnerable. SIRS, severe
shock, and organ failure may occur without necrosis or infection. Mortality
in the early phase from multisystem organ dysfunction and circulatory col-
lapse may be as high as 50% in patients with multisystem organ failure. The
late phase is less clearly defined and may persist for months after resolution
of the inflammatory cascade. The late phase involves the evolution, pro-
gression, and treatment of local complications of SAP. Mortality in the late
phase is primarily a result of sepsis-related organ dysfunction either from
infected necrosis or from infection in other organ systems of the debilitated
patient.
Volume Resuscitation/ICU Monitoring

Fluid resuscitation is the most important component of therapy for the
early phase of SAP. Resuscitation should be guided primarily by end-organ
perfusion (i.e., urine output) as with mild AP. Lactated Ringer solution is
the isotonic crystalloid fluid of choice for resuscitation. Patients with evi-
dence of severe disease require closer monitoring in an intensive care set-
ting. Central venous and arterial catheters are helpful guides for following
hemodynamic and volume status in these critically ill patients. Abdominal
compartment syndrome is a concern in SAP patients; in our robust clinical
experience, this condition is rarely seen.
Antibiotics
The second peak of mortality in SAP patients is almost ubiquitously due to
infection. The practice of prophylactic antibiotic administration emerged
in hopes of attenuating this second mortality peak. To date, 14 prospec-
tive randomized trials have compared antibiotic prophylaxis versus none
in SAP. Though all of these studies are underpowered and all suffer some
methodologic limitations, NO STUDY to date has definitively shown that
prophylactic antibiotic treatment affects SAP mortality. Therefore, broad-
spectrum antibiotic prophylaxis is not indicated with SAP.
Documented infections (bloodstream, urinary, pulmonary, etc.)
should be treated with discrete end point of antibiotic therapy. When infec-
tion is documented in (peri)pancreatic necrosis, broad-spectrum anti-
biotic treatment should be administered to provide coverage of GI flora.
Carbapenems are recommended for initial broad-spectrum coverage, but
therapy should be tailored according to local resistance patterns and indi-
vidual culture and susceptibility data.
Nutrition
Nutritional support should be initiated as soon as the patient is resus-
citated and hemodynamically stable. Enteral administration is prefer-
able when possible. Prospective data support oral or nasogastric feeding,
though many SAP patients will manifest gastric ileus and will tolerate
post-Treitz ligament feedings more comfortably. Parenteral nutrition is
002164307.INDD 20 7/14/2014 5:04:00 PM

often necessary to supplement caloric needs. Feeding gastrojejunostomy

or jejunostomy tube placement is appropriate if more than 30 days of
nutritional support are expected. Reinitiation of oral feeding is appro-
priate when abdominal pain improves, and many patients in the “hold-
ing pattern” of the second phase of SAP will derive nutrition from some
combination of oral, enteral, and parenteral nutrition. It is important to
remember that all SAP patients will remain catabolic until the inflamma-
tory focus has resolved.
Indications for ERCP

Recent meta-analysis of seven prospective, randomized trials found that
routine ERCP does not affect outcome in AP. Therefore, ERCP should be
reserved for those patients with biliary obstruction or cholangitis. Early
consultation with endoscopists is appropriate for patients with SAP.
Venous Thromboembolism
Venous thromboembolism incidence is remarkably common (>50%) in AP
patients, commonly affecting splanchnic vessels (splenic vein, superior
mesenteric vein, and portal vein) in addition to extremity veins. Venous
thrombosis may be associated with vascular catheterization. In general, we
do not routinely anticoagulate patients with splanchnic thrombosis (these
will usually not resolve until the underlying inflammatory focus and mass
effect from adjacent collections have resolved). Peripheral deep vein throm-
bosis, however, should be treated with anticoagulation. In addition, screen-
ing for peripheral DVT seems warranted in all patients with SAP/NP.
Bleeding
Bleeding severe AP/NP may be due to disruption of retroperitoneal veins
(which is almost always self-limiting) or from visceral arterial pseudoan-
eurysm (PSA). In NP patients, PSA may present with sudden increase of
abdominal pain, GI hemorrhage, or blood in a surgical or percutaneously
placed drain. CT angiography is an excellent first-line test with which to
diagnose (or exclude the presence of) PSA. Treatment of PSA is by transar-
terial embolization, which is successful in virtually all patients.
Ischemic Viscera—Colon and Gallbladder

Ischemia of the colon or gallbladder should be considered in any patient
with SAP who suddenly decompensates. Colon ischemia may occur in up
to 8% of SAP patients; the mechanism of colonic ischemia is likely related
to venous occlusion with subsequent tissue congestion. Unfortunately, the
only method to securely diagnose (or rule out) colon ischemia is by lapa-
rotomy. Ischemic cholecystitis may be treated initially with tube cholecys-
tostomy; however, these patients should be followed very closely with a low
threshold for operative exploration in those who do not demonstrate rapid
improvement after cholecystostomy tube placement.
Fine Needle Aspiration

Fine needle aspiration (FNA) may diagnose infected (peri)pancreatic
necrosis; however, no indication currently exists for routine FNA. Clinicians
should be aware of the significant (12% to 25%) false-negative rate associ-
ated with FNA.
Intervention in Necrotizing Pancreatitis

Infected pancreatic necrosis is nearly always mandates intervention to
achieve source control. The management of sterile necrosis is more contro-
versial. Clearly, some patients with symptomatic sterile necrosis will benefit
0002164307.INDD 21 7/14/2014 5:04:00 PM

from debridement. In addition, a number of patients with presumed sterile

necrosis will harbor occult infection. Experience has shown that definitive
intervention should not be undertaken earlier than 4 weeks from the dis-
ease onset. This time frame allows (peri)pancreatic collections to mature
and wall off, making debridement safer and easier to achieve at a single
setting. Earlier operation results in incomplete debridement of immature
necrosis. In addition, early operation is fraught with hazard and potentially
catastrophic bleeding complication.
The goals of treating NP include (1) safe debridement of all solid
necrotic material, (2) drainage of any pancreatic fistula (externally or if pos-
sible internally), (3) gaining access to the alimentary tract, and (4) cholecys-
tectomy (if technically possible and safe) in patients with biliary etiology.
The classic approach to treating NP has been open operative debridement.
Recently, enthusiasm for minimally invasive approaches to NP patients has
grown. These minimally invasive approaches include percutaneous drain-
age, endoscopic debridement, a combination of percutaneous/endoscopic
approach, retroperitoneal debridement (sinus tract necrosectomy or vid-
eoscopic assisted retroperitoneal debridement), laparoscopic transabdom-
inal debridement, and transgastric debridement (open or laparoscopic).
Regardless of the approach chosen, one physician must be willing to accept
responsibility for the duration of the NP patient’s care in what is commonly
a long-term (months to even years) recuperation.
Patient Selection for Intervention

NP is a very heterogeneous disease. It is critical for the practitioner to real-
ize that one approach does NOT fit all patients. Individual patients must be
approached on a case-by-case basis, ideally in the context of an interested
multidisciplinary group that includes pancreatic surgeons, therapeutic
endoscopists, and interventional radiologists. The appropriate interven-
tion in SAP/NP depends principally upon the location of the peripancreatic
collections and the volume of solid necrosis present. An important consid-
eration is the presence of pancreatic parenchymal necrosis, particularly
when this is associated with a major pancreatic duct disruption (Fig. 2.4).
This common finding will lead to the so-called disconnected pancreatic
duct syndrome (DPDS), which will predictably result in persistent pancre-
atic fistula when drained externally. Figure 2.5 depicts common patterns
of necrosis.
Antibiotic Treatment Alone

Several small and one moderately larger series have shown efficacy of anti-
biotic treatment alone in highly select patients with pancreatic necrosis.
This treatment strategy must be reserved for highly select patients, who
should be kept under very close follow-up. Mechanical intervention should
be applied for any sign of clinical deterioration.
FIGURE 2.4 CT of a patient with DPDS. Viable head (left panel, arrow) and tail (right
panel, arrow) are present; neck and body are necrotic, and subsequent drainage from the
tail is consolidated into a large lesser sac collection.
002164307.INDD 22 7/14/2014 5:04:01 PM

FIGURE 2.5 Patterns of necrosis. Left image (collection confined to lesser sac) may be most
amenable to transgastric approach. Middle image (collection extending down left para-
colic gutter) may be most amenable to retroperitoneal debridement. Right image (necrosis
involving left paracolic gutter, pancreatic head, and/or root of the small bowel mesentery)
is a very challenging situation and may be best approached by open debridement.
Percutaneous Drainage
Recent reports suggest that as many as one-third of all NP patients may
be successfully treated with percutaneous drainage alone. Importantly, this
treatment strategy mandates frequent (72-hour) repeat cross-sectional
imaging with drain upsizing and/or additional drain placement. Patients
with DPDS, pancreatic head necrosis, or necrosis that extends down the
root of the small bowel mesentery are not good candidates for the per-
cutaneous approach. Patients who develop infected necrosis prior to the
4-week time point should have percutaneous drains placed. Percutaneous
drainage in the setting of early infection often successfully temporizes the
clinical situation until safe (i.e., 4 weeks) definitive debridement may be
undertaken. If a patient does not rapidly improve after percutaneous drain-
age, the surgeon should consider laparotomy to exclude ischemic viscera as
a source of deterioration.
Endoscopic Necrosectomy
Transgastric endoscopic debridement of NP was first described in 1996.
This approach is attractive for its minimally invasive nature, as well as for
the concept of providing durable internal drainage for patients with DPDS.
Endoscopic debridement requires a dedicated endoscopist with advanced
procedural skills. Patients treated with this approach commonly require
multiple (4+) endoscopy sessions, each under general anesthetic.
Combined Endoscopic and Percutaneous Approach

The Virginia Mason group in Seattle has championed a combination pro-
cedure including endoscopic transgastric and percutaneous drainage. This
approach permits irrigation of the percutaneous drain, which may facilitate
mechanical egress of solid material into the alimentary tract. Dedicated cli-
nicians (both endoscopists and interventional radiologists) must follow the
patient treated with this combined approach, as they also require multiple
interventions.
0002164307.INDD 23 7/14/2014 5:04:02 PM

Retroperitoneal Approach
Select patients with appropriate anatomy may be debrided through the ret-
roperitoneum. The “step-up” approach of percutaneous drainage followed
at short interval by retroperitoneal debridement if necessary compared
favorably to direct open necrosectomy in the Dutch prospective, random-
ized PANTER trial. A multicenter North American study and another larger
series from Liverpool have also shown the efficacy of this approach in
select patients. Of particular interest is the finding that nearly one-third of
patients treated by the “step-up” approach resolved symptoms with percu-
taneous drainage alone. Further investigation is focusing on defining spe-
cific factors predicting success.
The etiology of AP must not be overlooked—minimally invasive
approaches such as percutaneous, endoscopic, and retroperitoneal do not
provide direct access for cholecystectomy.
Laparoscopic Transabdominal Debridement

Several small reports have documented the feasibility of laparoscopic (and
hand-assisted laparoscopic) debridement of peripancreatic and pancre-
atic necrosis, again in very select patients. Basic principles of pancreatic
debridement—that is, proper timing, thorough debridement, wide pan-
creatic drainage, and addressing underlying etiology (gallbladder) must
be followed in this approach. The totally laparoscopic approach is chal-
lenging due to lack of haptic feedback. General advantages of minimally
invasive surgery (i.e., shorter hospital stay, earlier return to work, etc.) may
be difficult to realize in this challenging group of patients.
Transgastric Debridement
Transgastric debridement can be accomplished endoscopically or surgi-
cally and is ideal for collections that are confined to the lesser sac and
for those patients with DPDS. Surgical transgastric debridement may be
approached laparoscopically or through a short upper midline incision.
Intraoperative ultrasound localizes the retrogastric collection; ultra-
sound is particularly helpful in patients with predominantly solid necro-
sis. Anterior gastrotomy is created, posterior gastrotomy is placed with
ultrasound guidance, and blunt necrosectomy performed through the
back wall of the stomach. Many NP patients have splenic vein thrombo-
sis with left-sided (sinistral) portal hypertension; gastric varices potenti-
ate significant hemorrhage in this situation. Long posterior gastrotomy is
incorporated into what essentially amounts to cyst-gastrostomy. Feeding
tube placement and cholecystectomy may be performed at the same set-
ting, if indicated.
Several reports of surgical transgastric debridement have been pub-
lished recently. This approach appears to be an excellent choice for select
patients with necrosis confined to the lesser sac but may not be as useful
for those with necrosis extending down the paracolic gutters or the small
bowel mesentery root. Long-term follow-up is accruing; patients should be
counseled regarding the potential for reaccumulation of retroperitoneal
collections or the potential for recurrent AP.
Open Debridement
Open pancreatic debridement represents the time-tested standard as a
safe, effective method to achieve thorough debridement of necrotic tis-
sue, wide drainage of the pancreas and debridement bed, access to the
alimentary tract, and cholecystectomy (if indicated by biliary etiology).
Contemporary outcomes of open debridement at experienced centers doc-
ument improved morbidity and mortality rates compared to earlier eras,
002164307.INDD 24 7/14/2014 5:04:02 PM

despite the fact that open surgical debridement is being reserved for those
patients with the most severe pathology.
Generous exposure should be obtained; the inflammatory response
is generally quite dense and mild venous oozing seen during mobiliza-
tion of tissue is usually self-limited. Intraoperative ultrasound comple-
ments preoperative imaging as a surgical “road map” to ensure thorough
debridement. Collections in the paracolic gutters provide a safe route by
which to access contiguous lesser sac collections. Special care should be
taken debriding necrosis around the pancreatic head; catastrophic hem-
orrhage may accompany inadvertent injury to the superior mesenteric or
portal veins. The vascular walls are commonly weakened by the surround-
ing inflammatory mass, and proximal/distal vascular control is extremely
challenging in this densely inflamed operative field. Gentle debridement
should only take tissue that is easily dislodged from the retroperitoneum.
Wide drainage of the necrosis bed is important to control pancreatic fistula
externally. In patients with biliary AP, the decision to perform simultane-
ous cholecystectomy should be based on the patient’s clinical condition; if
cholecystectomy is deferred, it should be performed in the near future (as
the patient recuperates) to avoid recurrent biliary problems or recurrent
pancreatitis. Enteral access is routine in patients requiring open pancreatic
debridement; our preference is to place a gastrojejunostomy feeding tube.
OUTCOMES
Most patients with mild AP will recuperate without major clinical conse-
quence. Up to 20% of patients with a first episode of AP will have recur-
rent AP, some of whom will progress to chronic pancreatitis. Patients with
NP, particularly those requiring debridement, often take several months
to return to baseline quality of life. Few data exist to document the true
incidence of long-term NP complications, which in addition to recurrent
AP may include exocrine and endocrine insufficiency, biliary stricture, duo-
denal stricture (particularly in patients with head necrosis), disconnected
pancreatic duct with pancreatic fistula, pseudocyst, or recurrent retro-
peritoneal collections/abscess. A pragmatic approach to long-term care
includes educating the patient as well as their primary care providers to the
possibility of long-term complication.
CONCLUSION
AP is a common, serious medical condition causing significant morbidity
and mortality. Surgeons caring for AP patients should be prepared to lead a
multidisciplinary team and be prepared to care for these patients over the
long term of their illness. The challenges of caring for this complex disease
process are rewarded by seeing the patients return to good life quality after
this debilitating, potentially fatal illness. Research efforts are focused on
understanding the disease biology in order to identify specific treatment.
Suggested Readings
Balthazar EJ, Robinson DL, Megibow AJ, et al. Acute pancreatitis: value of CT in estab-
lishing prognosis. Radiology 1990;174:331–336.
Banks PA, Bollen TL, Dervenis C, et al.,; Acute Pancreatitis Classification Working
Group. Classification of acute pancreatitis—2012: revision of the Atlanta classifi-
cation and definitions by international consensus. Gut 2013;62(1):102–111.
Banks PA, Gerzof SG, Langevin RE, et al. CT-guided aspiration of suspected pancreatic
infection: bacteriology and clinical outcome. Int J Pancreatol 1995;18:265–270.
Büchler MW, Gloor B, Müller CA, et al. Acute necrotizing pancreatitis: treatment strat-
egy according to the status of infection. Ann Surg 2000;232:619–626.
0002164307.INDD 25 7/14/2014 5:04:02 PM

Dellinger EP, Forsmark CE, Layer P, et al. Determinant-based classification of acute

pancreatitis severity. An international multidisciplinary consultation. Ann Surg
2012;256:875–880.
Petrov MS, Shanbhag S, Chakraborty M, et al. Organ failure and infection of pancre-
atic necrosis as determinants of mortality in patients with acute pancreatitis.
Gastroenterology 2010;139:813–820.
van Santvoort HC, Besselink MG, Bakker OJ, et al. A step-up approach or open
necrosectomy for necrotizing pancreatitis (PANTER trial). N Engl J Med
2010;362:1491–1502.
002164307.INDD 26 7/14/2014 5:04:02 PM

3 Chronic Pancreatitis
Katherine A. Morgan
Adventures are all very well in their place, but there’s a lot to be said for regular
meals and freedom from pain
Neil Gaiman
INTRODUCTION
Chronic pancreatitis (CP) is a vexing disorder, marked clinically by
intractable debilitating abdominal pain and attendant nutritional failure.
CP is complex in its pathogenesis, clinical management, and psychoso-
cial implications. Thus, management challenges arise on all levels in the
endeavor to understand and treat this difficult disease.
EPIDEMIOLOGY
CP is a significant public health care concern. The incidence and preva-
lence of CP in the United States have been estimated to be 4/100,000 and
41/100,000, respectively. Mortality from CP is increased up to 3.5 times the
average population, with a 10-year survival approximating 50%. US health
care costs for pancreatitis were estimated at $3.7 billion in 2009.
PATHOGENESIS
The etiology and pathophysiology underlying CP are poorly understood.
Pancreatitis is likely a heterogenous grouping of different diseases, mani-
festing as a similar end histologic result of pancreatic inflammation and
fibrosis. Risk factors, most evidently excessive alcohol consumption, have
been implicated in disease pathogenesis. More likely, however, an intricate
interplay of environmental factors, genetic susceptibility, and host immu-
nologic response results in disease development.
Several theories on the etiology of CP have been put forth. The necro-
sis–fibrosis hypothesis holds that repeated bouts of acute pancreatitis
result in organ injury, resulting in fibrosis and CP. A similar model is the
sentinel acute pancreatitis event hypothesis, where an environmental
stress in a susceptible host leads to an inflammatory response (acute pan-
creatitis), which then incites a host-specific immune response resulting in
fibrosis (CP).
Alcohol causes oxidative stress and mitochondrial damage, and it
lowers the threshold for trypsinogen activation. It also shifts cell death
from apoptosis to necrosis and stimulates immune-related inflammatory
fibrosis. While excessive alcohol consumption is liable in alcoholic pancre-
atitis, it is not a lone causative agent, as evidenced by the fact that most
alcoholics do not develop pancreatitis. Clearly, there is an underlying host
susceptibility factor involved. Similarly, tobacco is implicated as a risk
factor for CP, with increasingly recognized importance.
27
0002086372.INDD 27 7/14/2014 2:34:57 PM

TABLE
3.1 Genes Implicated in Chronic Pancreatitis
Gene Mechanism
Cationic trypsinogen gene (PRSS1) Inappropriate activation of trypsin
Serum protease inhibitor, Kazal type Failed inhibition of activated trypsin
1 (SPINK 1)
Cystic fibrosis transmembrane Decreased pH and flow of pancreatic
regulator (CFTR) juice
Chymotrypsin C (CTRC) Failed inhibition of prematurely
activated trypsin
Calcium sensing receptor (CaSR) Failed intraductal calcium
homeostasis
The past two decades have brought revolutionary understand-

ing to the paradigm of pancreatitis pathophysiology with the discovery
of the cationic trypsinogen gene (PRSS1) and the delineation of heredi-
tary pancreatitis in 1996. This important revelation clearly demonstrates
the phenotypic result of alterations in the trypsinogen pathway, with an
autosomal dominant disorder resulting in potentially severe disease with
increased susceptibility for cancer development. It has become a milestone
model in the genetic basis of pancreatic disease. Since then, several other
genes have been implicated in the pathogenesis of pancreatitis, all involv-
ing the trypsinogen pathway, including SPINK 1 (serum protease inhibitor,
Kazal type 1), CFTR (cystic fibrosis transmembrane regulator), CTRC (chy-
motrypsin C), and CaSR (calcium-sensing receptor) (Table 3.1).
Important immunologic discoveries including correlations with vari-
ations in human leukocyte antigen antigens have come about in the past
decade. On a cellular level, pancreatic stellate cells are identified in their
important role of extracellular matrix formation and fibrosis development.
The understanding of the contribution of substance P, nerve growth factor,
brain-derived growth factor, and other cell signaling molecules to disease
pathogenesis is evolving. Elucidation of the mechanism of disease develop-
ment on the genetic and cellular level explicates the concept of individual
susceptibility in the face of environmental stressors.
CLINICAL PRESENTATION
Patients may present with complications of pancreatitis such as obstruction
(biliary, duodenal, mesenteric vascular thrombosis), perforation (pseudocyst,
pancreatic fistula, pancreatic ascites), or bleeding (visceral arterial pseudoa-
neurysm). The clinical hallmark of CP, however, is intractable, debilitating
abdominal pain. Pain is reported by 80% to 94% of patients with CP. It is typi-
cally described as sharp and burning, located in the epigastrium and radiating
around to the back. Nausea, emesis, and food intolerance often accompany the
pain. Patients may not have objective findings associated with pain exacerba-
tion episodes, such as serum amylase elevation, thus leading to challenges for
health care workers and patients alike. Commonly, patients with severe disease
are challenged in their relationships with family and health care workers and
may develop maladaptive behaviors and associated social marginalization.
Endocrine and exocrine organ failure develops with progressive disease.
Approximately 60% of CP patients will develop insulin-dependent diabetes.
Greater than 90% loss of exocrine function must occur to result in malab-
sorption, with absorption of fat more significantly affected than protein.
002086372.INDD 28 7/14/2014 2:34:57 PM

Chapter 3 / Chronic Pancreatitis 29
DIAGNOSIS
The diagnosis of CP entails a consistent clinical history, notable for
pancreatic-type abdominal pain, along with radiographic evidence of dis-
ease. The most relevant imaging modalities include contrast-enhanced
computed tomography (CT), secretin-stimulated magnetic resonance pan-
creatography (MRP), endoscopic retrograde pancreatography (ERP), and
endoscopic ultrasound (EUS).
Abdominal CT may show calcifications, parenchymal thickening or
atrophy, or pancreatic ductal dilation evidencing disease. Alterations in
parenchymal enhancement, decreased response to secretin stimulation,
and pancreatic ductal pathology such as strictures or dilation may be better
delineated by MRP. ERP can show ductal changes and historically has been
the gold standard of objective disease diagnosis, although MRP is quickly
replacing ERP for diagnostic purposes at many centers. The ERP-derived
Cambridge classification system is the standard disease grading system
derived from an international consensus (Table 3.2). EUS is touted as the
most sensitive of the imaging modalities, although its utility is limited by
interobserver variability. It, too, has a disease grading system based on the
identification of objective features of CP (Table 3.3).
MANAGEMENT
In CP patients with debilitating pain who have failed medical management
to include alcohol and tobacco abstinence, pain control, nutritional opti-
mization, and behavioral therapies, intervention is indicated for pain relief.
ENDOSCOPIC MANAGEMENT
Therapeutic ERP interventions for CP are undertaken with the goal of
relieving an obstructive process. Maneuvers include sphincterotomy,
stone extraction, stricture dilation, and stenting. In practice, the endo-
scopic approach is exhausted prior to surgery, given the perceived
advantages of a less invasive procedure, despite two randomized con-
trolled trials demonstrating improved outcomes with surgery in patients
with obstructive pancreatopathy. In the first trial by Dite and colleagues
from the Czech Republic, 72 patients were randomized to endoscopic or
TABLE
Cambridge Classification System for Severity of Chronic
3.2 Pancreatitis by ERCP Imaging
Grade Severity of Pancreatitis ERCP Findings

1 Normal No abnormal features
2 Equivocal Less than 3 abnormal branches
3 Mild More than 3 abnormal branches
4 Moderate Abnormal main duct and branches
5 Marked As above with one or more of:
Large cavities (> 10 mm)
Gross gland enlargement (> 2 × normal)
Intraductal filling defects or calculi
Duct obstruction, structure or gross
irregularity
Contiguous organ invasion
ERCP, endoscopic retrograde cholangiopancreatography
From Sarner M, Cotton PB. Classification of pancreatitis. Gut 1984;25:756–759.
0002086372.INDD 29 7/14/2014 2:34:57 PM

TABLE
3.3 Conventional EUS Criteria for Chronic Pancreatitis
Parenchymal Criteria Duct Criteria

Hyperechoic foci Irregular duct contour
Hyperechoic strands Visible side branches
Hyperechoic lobules, foci, or areas Hyperechoic duct margin
Cyst Dilated main duct
Normal (low probability) 0–2 criteria present
Indeterminate (intermediate 3–4 criteria present
probability)
High probability 5–9 criteria present
EUS, endoscopic ultrasound
From Wallace MB, Hawes RH, Durkalski V, et al. The reliability of EUS for the diagnosis for chronic
pancreatitis: interobserver agreement among experienced endosonographers. Gastrointest Endosc
2001;53:294–299.
s urgical intervention for pancreatic duct obstruction and associated pain.

Endoscopic therapy consisted of 52% sphincterotomy and stenting and 23%
stone removal. Operative management included 20% drainage procedures
and 80% resections. At 5-year follow-up, the surgical group had a greater
proportion of patients who were pain free (34% vs. 15%), while the rate of
partial pain relief was equivalent between the groups (52% surgery, 46%
endoscopy). In another trial by Cahen and colleagues from the Netherlands,
39 patients were randomized to endoscopic or surgical intervention for
dilated duct pancreatitis and pain. Endoscopic therapy included sphincter-
otomy and stent, while operative therapy was with a longitudinal pancre-
aticojejunostomy (LR-LPJ). At 5-year follow-up, pain relief was achieved in
80% of the surgical group versus 38% of the endoscopic group (p = 0.001).
The surgical group also had larger improvements in quality of life and
underwent fewer procedures. Equivalent morbidity, length of stay, and pan-
creatic function were seen in the two groups.
SURGICAL MANAGEMENT
The primary indication for surgical intervention in CP is intractable pain,
unresponsive to medical therapies. Forty percent to sixty-seven percent
of patients with CP will meet these criteria for surgery. The goals of sur-
gery are to effectively and durably relieve pain while minimizing morbidity
including preserving pancreatic parenchyma when possible. The etiology of
pancreatic pain is poorly understood and likely multifactorial. Thus, opera-
tive decision making can be challenging. The pancreatic ductal anatomy is
the primary determinant in surgical planning. Generally speaking, patients
with a large main pancreatic duct (> 6–7 mm in diameter) are presumed to
have an obstructive component to their disease and are thus well served by
a drainage procedure. In patients with small duct disease, resection of dam-
aged and poorly drained parenchyma is typically effective. In patients with
head-predominant or tail-centered disease, a directed resection is optimal.
In patients with a small pancreatic duct and diffuse organ involvement,
a total pancreatectomy (TP) with islet autotransplantation (IAT) may be
indicated.
The heterogenous nature of CP anatomical changes mandates the sur-
geon’s familiarity with several operative approaches—no one operation is
suitable for every patient.
002086372.INDD 30 7/14/2014 2:34:57 PM

Longitudinal Pancreaticojejunostomy
Operative pancreatic drainage for pancreatitis was described in a small
series of patients by Puestow and Gillesby in 1957. A modification of this
original drainage procedure that more closely resembles modern day
technique was reported by Partington and Rochelle in 1960. The clas-
sic operation for pancreatic drainage, the lateral pancreaticojejunostomy
(LPJ), entails opening the pancreatic duct anteriorly along its length within
a fibrotic gland, medially to the level of the gastroduodenal artery. The
opened pancreatic duct is then anastomosed to a Roux-en-Y jejunal limb
(Fig. 3.1).
Potential significant procedure-specific complications include intra-
operative hemorrhage due to splenic vein or gastroduodenal artery injury,
postoperative hemorrhage typically from the gastroduodenal artery, and
anastomotic leak seen in 10% of cases.
Multiple retrospective case series have been reported evaluating out-
comes with LPJ, with pain relief rates of 48% to 91%. Morbidity rates are low
(20% on average) and endocrine and exocrine function is often preserved.
LPJ is an effective and safe procedure for pain relief in many patients with
dilated duct pancreatitis. A secondary failure rate does exist with LPJ,
often attributed to disease burden in the head of the pancreas. Intraductal
stone disease in the head of the pancreas can be cleared with intraopera-
tive pancreatoscopy and lithotripsy, which has been shown to improve
outcomes (reduced readmissions, increased pain relief rates). Patients with
significant burden of inflammatory fibrosis in the head of the pancreas may
be better served with a localized pancreatic head resection along with a
pancreaticojejunostomy.
Local Resection of the Pancreatic Head with Longitudinal

Pancreaticojejunostomy
In the mid-1980s, Frey and colleagues described adding a localized resec-
tion of the head of the pancreas along with an LR-LPJ. This procedure
addresses the inflammatory disease burden often found in the head of the
pancreas in patients with dilated duct pancreatitis. In addition, LR-LPJ has
the advantage of duodenal preservation. In modern series, pain relief rates
of 62% to 88% are reported, with morbidity of 20% to 30% (Fig. 3.2).
Pancreatoduodenectomy
In 1946, Dr. Whipple described pancreatoduodenectomy (PD) for CP.
Patients with an inflammatory mass in the head of the pancreas and those
with diffuse small duct disease with the head as the suspected pacemaker
of disease may benefit from PD. Pain relief rates of 70% to 89%, morbid-
ity of 16% to 53%, and mortality of less than 5% are reported. Procedure-
specific significant complications include intraoperative hemorrhage from
portal vein injury, postoperative biliary leak, postoperative pancreatic fis-
tula (POPF), postoperative hemorrhage from the ligated gastroduodenal
artery, and delayed gastric emptying, the latter two often associated with
postoperative pancreatic leak. Grading of POPF is according to clinical sig-
nificance (Table 3.4).
The pylorus-preserving pancreatoduodenectomy (PPPD) was popular-
ized in the 1970s by Traverso and Longmire and has been embraced by many
pancreatic surgeons. Pylorus preservation is intended to improve nutritional
outcomes, although studies have not confirmed this proposed advantage.
Pain relief rates and endocrine outcomes are similar between the classic PD
(with antrectomy) and PPPD, although improved professional rehabilitation
and improved quality of life after PPPD have been touted. Technical conduct
of PD (and distal pancreatectomy [DP]) is detailed in Chapter 7.
0002086372.INDD 31 7/14/2014 2:34:57 PM

Aspiration of dilated
pancreatic duct
A
B
FIGURE 3.1 A. After exposure of the full length of the pancreas, the pancreatic duct is identi-
fied by aspiration using a 21-gauge needle on a 5-mL syringe. B. The Roux-en-Y limb is brought
alongside the pancreas in an isoperistaltic fashion. The longitudinal pancreatic ductotomy is vis-
ible. Two corner stitches are placed between the jejunum and the apices of the ductotomy. An
enterotomy is then made lengthwise along the jejunum, in parallel to the pancreatic ductotomy.
(Continued)
002086372.INDD 32 7/14/2014 2:35:04 PM

Omentum
Stomach Pancreas
Duct
Duodenum
Roux-en-Y
jejunal loop
Transverse Transverse
colon mesocolon
Stomach
Pancreas
Duodenum
Roux-en-Y
jejunal loop
C
FIGURE 3.1 (Continued) C. A completed Roux-en-Y lateral pancreatojejunostomy. (From
Lillemoe K, Jarnigan W, eds. Master techniques in surgery: hepatobiliary and pancreatic
surgery. Philadelphia, PA: Lippincott Williams & Wilkins, 2013.)
Duodenum-Preserving Pancreatic Head Resection

In the 1970s, Beger described the technique of duodenum-preserving pan-
creatic head resection. In this procedure, the inflamed and fibrotic pancre-
atic head tissue is resected, leaving the duodenum and terminal bile duct
intact. The intent is to minimize surgical morbidity, including the long-term
morbidity of duodenal resection. Pain relief after this procedure has been
reported in 77% to 88% of patients with professional rehabilitation in 63% to
69% and morbidity of 29%. Several variations of duodenum-preserving pan-
creatic head resection have evolved at various centers, primarily in Europe
(Fig. 3.3).
Multiple randomized controlled trials comparing the different
approaches to pancreatic head resection have been undertaken and are
summarized in Table 3.5. While the advantage of any single procedure has
not been demonstrated, this effort in outcome research has been important
in moving the field of pancreatic surgery forward.
0002086372.INDD 33 7/14/2014 2:35:09 PM

B
FIGURE 3.2 Frey procedure. A. The Frey procedure combines a circumscript excision
in the pancreatic head with longitudinal opening of the pancreatic duct toward the tail.
B. Reconstruction is performed with an anastomosis with a Roux-en-Y jejunal loop.
Compared to the Beger procedure, the extent of resection of the pancreatic head is smaller;
however, reconstruction is easier as it only requires one anastomosis to the pancreas. (From
Lillemoe K, Jarnigan W, eds. Master techniques in surgery: hepatobiliary and pancreatic
002086372.INDD 34 7/14/2014 2:35:12 PM

TABLE
3.4 Grading System for Postoperative Pancreatic Fistula
Grade A B C
Clinical condition Well Often well Ill appearing/bad
Specific treatment No Yes/no Yes
US/CT (if obtained) Negative Negative/positive Positive
Persistent drainage No Usually yes Yes
(after 3 wk)
Reoperation No No Yes
Death related to POPF No No Possibly yes
Signs of infection No Yes/no Yes
Sepsis No No Yes
Readmission No Yes/no Yes/no
US, ultrasound; CT, computed tomography; POPF, postoperative pancreatic fistula.
From Bassi C, Dervenis C, Butturini G, et al. Postoperative pancreatic fistula: an international study
group (ISGPF) definition. Surgery 2005;138:8–13.
Distal Pancreatectomy
In CP patients with a midpancreatic body ductal stricture or those with dis-
ease localized to the body and tail of the pancreas, a DP can be beneficial.
Pain relief rates of 57% to 84% are reported with return to work in 29% to
73% and morbidity rates of 15% to 32%. POPF is the most important poten-
tial procedure-specific complication and occurs in up to 30% of cases.
A
FIGURE 3.3 Beger procedure. A. The pancreas is dissected on the level of the portal vein.
The pancreatic head is excavated and the duodenum is preserved with a thin layer of pan-
creatic tissue. If the bile duct is obstructed, it can be opened and an internal anastomosis
with the excavated pancreatic head can be performed as shown.
(Continued)
0002086372.INDD 35 7/14/2014 2:35:16 PM

B
FIGURE 3.3 (Continued) B. The reconstruction is performed with two anastomoses, of
the pancreatic tail remnant and of the excavated pancreatic head with a Roux-en-Y jejunal
loop. (From Lillemoe K, Jarnigan W, eds. Master techniques in surgery: hepatobiliary and
pancreatic surgery. Philadelphia, PA: Lippincott Williams & Wilkins, 2013.)
Total Pancreatectomy with Islet Autotransplantation

An early description of TP for CP comes from Dr. Clagett in 1944. Tellingly,
his patient died 10 weeks postoperatively from a hypoglycemic event. In
some patients with debilitating pain from CP, TP can be an effective means
of pain relief. Candidates for TP include patients with diffuse small duct
pancreatitis, those that have failed lesser procedures, and those with hered-
itary pancreatitis.
Parenchymal preserving procedures are always considered preferen-
tially to TP when possible to minimize morbidity. The obligatory incipient
pancreatogenic diabetes that results from TP can be prohibitively morbid,
even with current diabetes management tools including the insulin pump.
The largest recent series of patients undergoing TP is from the Mayo Clinic
in 2005. In that series, 26% of long-term survivors were rehospitalized for
glycemic control, and diabetes-related quality of life was poor.
In the 1970s, IAT was developed at the University of Minnesota, where
islets are harvested from the resected pancreas immediately after resection
and returned to the patient through the portal vein into the liver. The first
successful procedure was described by Dr. Sutherland in 1978. Over the
002086372.INDD 36 7/14/2014 2:35:20 PM

TABLE
3.5
Comparative Randomized Controlled Trials of Pancreatic
Head Resection
Study Comparison N Morbidity Mortality Pain Relief, %

Klempa et al. 1995 PD 21 51 0 70
DPPHR 22 54 5 82
Buchler et al. 1995 PPPD 20 20 0 67
DPPHR 20 15 0 94
Farkas et al. 2006 PPPD 20 40 0 90
DPPHR 20 0 0 85
Izbicki et al. 1995 DPPHR 20 20 0 95
LR-LPJ 22 9 0 89
Izbicki et al. 1998 LR-LPJ 31 19 3 90
PPPD 30 53 0 71
Strate et al. 2005 DPPHR 38 NR NR 82
LR-LPJ 36 NR NR 81
PD, pancreatoduodenectomy; DPPHR, duodenum-preserving pancreatic head resection; PPPD,
pylorus-preserving pancreatic head resection; LR-LPJ, local resection pancreatic head with

longitudinal pancreaticojejunostomy.
past decade, interest in this technique has grown, with now several centers
offering TP–IAT for CP.
Morbidity rates of 48% to 56% and 0% to 2% mortality are reported
in TP–IAT. Pain relief is described in 80% to 85% patients after TP–IAT.
Statistically significant improvements in both physical and mental
health–related quality of life are noted as early as 6 months postopera-
tively and appear durable in 3-year follow-up. Insulin independence is
described in 25% to 40% patients in short-term follow-up up to 3 years.
Longer-term follow-up demonstrates decline in islet function over time;
however, C-peptide production has been documented up to 13 years post-
operatively. Longer-term follow-up is needed for this relatively radical
procedure.
Minimally Invasive Pancreatic Surgery

Laparoscopic pancreatic resection originated in the 1990s with the explo-
sion of laparoscopy and the associated development of laparoscopic tools
including the endoscopic stapler. Laparoscopic distal pancreatectomy
(LDP) is the most commonly performed laparoscopic pancreas resection. A
large multi-institutional study evaluated 667 DPs, 159 (24%) of which were
laparoscopic. Of these, only 14 (9%) of these LDPs were for CP. While LDP
is most commonly performed for benign or low-grade neoplasms, it is also
applied to patients with CP. The inflammatory distortion of anatomy, loss
of normal tissue planes, and frequent coexistence of sinistral (left-sided)
portal hypertension in CP make these cases extremely challenging.
Laparoscopic PD is being performed at many centers now through-
out the world. The largest single-center series is from India by Palanivelu
with 75 patients, with mean operative time of 357 minutes, blood loss
74 mL, hospital stay 8.2 days, morbidity 26.7%, and mortality 1.3%. At the
Mayo Clinic, Kendrick has reported on 65 patients, with outcomes similarly
comparable to open resection, and has additionally reported on 11 patients
safely undergoing major venous reconstruction via a totally laparoscopic
route. The limitations of laparoscopic PD in the CP patient are similar to
those encountered for DP.
0002086372.INDD 37 7/14/2014 2:35:20 PM

CONCLUSIONS
CP remains a challenging disorder in the current era. In basic science, progress
is occurring in the understanding of underlying genetic and cellular mecha-
nisms of disease, with hopes for smart pharmacologic interventions includ-
ing enzyme pathway inhibition. On the surgical front, sophisticated therapies
are evolving including minimally invasive techniques and TP with IAT. In
addition, in neuroscience, the recognition of neuroplasticity and maladaptive
central pain pathways as an important factor in CP pain management holds
promise for continued progress in the management of this difficult disease.
Suggested Readings
Bachem MG, Zhou Z, Zhou S, et al. Role of stellate cells in pancreatic fibrogenesis associ-
ated with acute and chronic pancreatitis. J Gastroenterol Hepatol 2006;21:S92–S96.
Beger HG, Krautzberger W, Bittner R, et al. Duodenum preserving resection of the
head of the pancreas in severe chronic pancreatitis. Surgery 1985;97:465–475.
Buchler MW, Friess H, Muller MW, et al. Randomized trial of duodenum preserving
pancreatic head resection versus pylorus preserving whipple in chronic pancre-
atitis. Am J Surg 1995;169:65–69.
Cahen DL, Gouma DJ, Laramee P, et al. Long-term outcomes of endoscopic vs
surgical drainage of the pancreatic duct in patients with chronic pancreatitis.
Gastroenterology 2011;141:1690–1695.
Demir IE, Teiftrunk E, Maak M, et al. Pain mechanisms in chronic pancreatitis: of a
master and his fire. Langenbecks Arch Surg 2011;396:151–160.
DiMagno EP, Go VL, Summerskill WH. Relations between pancreatic enzyme outputs
and malabsorption in severe pancreatic insufficiency. N Engl J Med 1973;288:813–815.
Farkas G, Leindler L, Daroczi M, Farkas G. Prospective randomized comparison of
organ preserving pancreatic head resection with pylorus preserving pancreato-
duodenectomy. Langenbecks Arch Surg 2006;391:338–342.
Frey CF, Smith GJ. Description and rationale of a new operation for chronic pancreati-
tis. Pancreas 1987;2:701–707.
Gorry MC, Gabbaizedah D, Furey W, et al. Mutations in the cationic trypsinogen gene
are associated with recurrent acute and chronic pancreatitis. Gastroenterology
1997;113:1063–1068.
Izbicki JR, Bloechle C, Broering DC, et al. Extended drainage versus resection in sur-
gery for chronic pancreatitis: A prospective randomized trial comparing the lon-
gitudinal pancreaticojejunostomy combined with local pancreatic head excision
with the pylorus preserving pancreatoduodenectomy. Ann Surg 1998;228:771–779.
Izbicki JR, Bloechle C, Knoefel WT, et al. Duodenum preserving resection of head of
the pancreas in chronic pancreatitis. A prospective randomized trial. Ann Surg
1995;221:350–358.
Kendrick ML, Cusati D. Total laparoscopic pancreaticoduodenectomy: feasibility and
outcome in an early experience. Arch Surg 2010;145:19–23.
Klempa I, Spatny M, Menzel J, et al. Pancreatic function and quality of life after resec-
tion of the head of the pancreas in chronic pancreatitis. A prospective, random-
ized comparative study after duodenal preserving resection of the head of the
pancreas versus Whipple’s operation. Chirurg 1995;66:350–359.
Morgan K, Owczarski SM, Borckardt J, et al. Pain control and quality of life after pan-
createctomy with islet autotransplantation for chronic pancreatitis. J Gastrointest
Surg 2012;16:129–134.
Nealon WH, Thompson JC. Progressive loss of pancreatic function in chronic pancre-
atitis is delayed by main pancreatic duct decompression. A longitudinal prospec-
tive analysis of the modified Puestow procedure. Ann Surg 1993;217:458–468.
Schnelldorfer T, Mauldin PD, Lewin DN, et al. Distal pancreatectomy for chronic
pancreatitis. Risk factors for postoperative pancreatic fistula. J Gastrointest Surg
2007;11:991–997.
Strate T, Taherpour Z, Bloechle C, et al. Long term follow up of a randomized trial
comparing the Beger and Frey procedures for patients suffering from chronic pan-
creatitis. Ann Surg 2005;241:591–598.
Yadav D, Timmons L, Benson JT, et al. Incidence, prevalence, and survival of chronic
pancreatitis: a population based study. Am J Gastroenterol 2011;106:2192–2199.
002086372.INDD 38 7/14/2014 2:35:20 PM

4
Pancreatic Adenocarcinoma
Jashodeep Datta, Jeffrey A. Drebin,
and Charles M. Vollmer Jr
EPIDEMIOLOGY
Cancer of the exocrine pancreas, commonly referred to as pancreatic ductal
adenocarcinoma (PDAC), affects over 44,000 people a year and is the fourth
leading cause of cancer-related deaths in the United States with approxi-
mately 37,000 annual mortalities. Approximately 11 new cases of PDAC
are diagnosed per 100,000 of the population annually. At diagnosis, most
patients are advanced (metastatic), with only 15% to 20% of patients being
candidates for a potentially curative surgical resection. Resection offers the
best possibility for cure (particularly when margin negative and node nega-
tive), with 5-year survival rates around 20% when performed at specialized
high-volume centers. Nevertheless, median survival after resection is only
about 2 years and, for all stages, 5-year survival is an abysmal 3%.
PATHOLOGY
Solid Epithelial Neoplasms of the Exocrine Pancreas
The most common solid exocrine pancreatic tumors are pancreatic ductal
adenocarcinoma and its variants, acinar cell carcinoma and pancreatoblas-
toma. PDAC accounts for greater than 90% of all exocrine tumors of the
pancreas and represents the vast majority of cases referred for surgical con-
sultation. Thus, the rest of the discussion in this chapter will focus on the
management of PDAC. Histologically, the neoplastic cells in PDAC show
ductal differentiation, induce an intense and characteristic desmoplastic
stromal reaction, and display an infiltrative growth pattern. PDAC is thought
to develop from benign proliferative precursor lesions, termed pancreatic
intraepithelial neoplasia (PanIN), as a result of the progressive accumulation
of tumorigenic mutations. Variants of PDAC include giant cell, adenosqua-
mous, mucinous (noncystic), and anaplastic carcinomas. Acinar cell carci-
nomas tend to form larger tumors than PDAC, form acini which display an
eosinophilic and granular cytoplasm, and have a slightly better prognosis
than PDAC. Pancreatoblastomas occur in children 1 to 15 years old, con-
tain both epithelial and mesenchymal components, are usually larger than
10 cm, and have a more favorable prognosis than PDAC.
ETIOLOGY
Risk Factors for Pancreatic Ductal Adenocarcinoma
The risk of developing PDAC is related to demographic, acquired, and host
factors. Overall, any given person has a 0.5% risk of developing PDAC by
age 70.
(a) Demographic Factors: The peak incidence of PDAC is in the seventh to
eighth decades, with more than 80% of cases occurring between 60 and
80 years. PDAC is more common in men compared to females, with a rela-
tive risk of 1.35. Compared with Whites, both the incidence and mortality
39
0002086373.INDD 39 7/14/2014 2:43:49 PM

rates for PDAC are higher in African Americans of both genders. Other
demographic characteristics associated with PDAC include lower socio-
economic status, migrant status, and Ashkenazi Jewish heritage.
(b) Acquired or Environmental Factors: Tobacco smoke exposure is the most
consistently observed environmental risk factor for the development of
PDAC, while the contributory role of alcohol or various dietary substances
is more equivocal. PDAC risk increases in obese patients—a large cohort
study showed a 1.72 relative risk of PDAC in patients with body mass
index (BMI) ≥ 30 kg/m2 compared to individuals with BMI ≤ 23 kg/m2.
Moreover, moderate physical activity was associated with decreased
PDAC rates. An association between diabetes mellitus (DM) and PDAC
has also been posited due to the observation that approximately 80% of
PDAC patients have impaired glucose metabolism, impaired glucose
tolerance, or DM. Type 2 DM of at least 5 years’ duration has been
demonstrated to increase the risk of PDAC twofold. However, among
newly diagnosed diabetics, only 0.85% developed PDAC within 3 years.
Additionally, chronic pancreatitis, cystic fibrosis, benign endocrine
tumors, and pernicious anemia have also been implicated in elevating
PDAC risk.
(c) Host Factors: Only around 2% of PDAC can be considered “inherited.”
This consists of two distinct cohorts: (i) familial lineage of PDAC (genet-
ics as yet undetermined) and (ii) genetic syndromes caused by recog-
nized germ line mutations. These are as follows: (i) hereditary pancreatitis
(50-fold risk): autosomal dominant, PRSSI mutation on gene locus 7q35,
and recurrent pancreatitis from a young age; (ii) hereditary nonpolyposis
colorectal cancer (HNPCC type II) (eightfold risk): autosomal dominant,
MSH2/MLH1 mutation, and colonic, endometrial, and gastric cancers;
(iii) Peutz-Jeghers syndrome (32-fold risk): autosomal dominant, STK11/
LKB1 mutation on 19p13, and gastrointestinal hamartomatous polyps
and mucocutaneous pigmentation; (iv) familial atypical multiple mole
melanoma (FAMMM) syndrome (12- to 20-fold risk): autosomal domi-
nant, p16 mutation on 9p21, and multiple atypical nevi and melanoma;
(v) hereditary breast and ovarian cancer (5- to 10-fold risk): autosomal
dominant, BRCA2 mutation on 13q12–13, and breast and ovarian can-
cers; and (vi) ataxia–telangiectasia: autosomal recessive, ATM mutation
on 11q22–23, and ataxia, telangiectasia, and hematologic malignan-
cies. In addition, analysis of kindreds enrolled in the National Familial
Pancreas Tumor Registry revealed an 18-fold increased risk of PDAC in
familial PDAC kindreds compared to sporadic PDAC groups. The germ
line mutation(s) responsible for the familial predisposition to PDAC is/
are yet to be elaborated. Elevated risk begins with at least two affected
first-degree relatives and escalates significantly with each additional
family member. There is no known established relationship between a
family history of pancreatic cancer and the development of pancreatic
cystic neoplasms that devolve into PDAC.
Molecular Genetics
Significant advances have been made in understanding the molecular
genetics associated with PDAC in recent years. The pathogenesis of PDAC
involves alterations in the following: (i) tumor suppressor genes: p16, p53,
MADH4/DPC4, and BRCA2 are found to be inactivated in up to 90%, 75%,
55%, and 7% of sporadic PDAC cases, respectively; (ii) oncogenes: activating
point mutations in K-ras—a guanine nucleotide–binding protein involved
in growth factor signal transduction—are seen in between 80% and 100%
of PDAC cases. In spite of being widely investigated as a potential diagnos-
tic or prognosticative marker in PDAC, K-ras cannot be recommended for
002086373.INDD 40 7/14/2014 2:43:49 PM

Chapter 4 / Pancreatic Adenocarcinoma 41
PDAC detection based on the available data. Other oncogenes implicated

in PDAC pathogenesis are BRAF, AKT2, and MYB; and (iii) DNA mismatch
repair genes: approximately 4% of PDAC appears related to dysregulation of
DNA mismatch repair genes—MSH2, MLH1, PMS1, PMS2, MSH6/GTBP, and
MSH3. It is postulated that these mutations occur in a cumulative fashion
over the life span of a developing PDAC.
DIAGNOSIS
Clinical Presentation
The nonspecific symptoms associated with early PDAC—anorexia, nau-
sea, and weight loss—typically preclude its diagnosis at an early stage in
most cases. The majority of patients with right-sided PDAC present with
jaundice secondary to obstruction of the intrapancreatic portion of the
common bile duct and subsequently develop dark urine, light-colored
stools, and pruritus. These tumors can also cause mechanical obstruction
of the duodenal C-loop, leading to weight loss, gastric outlet obstruction
and, ultimately, vomiting. Left-sided tumors generally do not cause jaun-
dice and may present with epigastric/back pain from celiac axis lympho-
vascular infiltration or focal, segmental pancreatic duct obstruction or,
sometimes, duodenal obstruction at the ligament of Treitz. More subtle
presentations may be encountered in elderly patients with new-onset DM
or with episodes of acute pancreatitis in the absence of cholelithiasis or
ethanol abuse. The most consistent physical finding in PDAC is jaundice,
seen in up to 87% in head-based tumors but just 13% of distal (usually
body) tumors. Hepatomegaly, palpable gallbladder (Courvoisier sign), and
malnutrition may be observed. Signs of advanced disease include ascites,
cachexia, liver nodularity, left supraclavicular adenopathy (Virchow node),
periumbilical adenopathy (Sister Mary Joseph node), and pelvic drop
metastasis (Blumer shelf).
Laboratory Analysis
Workup usually reveals an elevated total and conjugated serum biliru-
bin, alkaline phosphatase, and γ-glutamyl transferase reflective of biliary
obstruction. Transaminases, amylase, and lipase may be normal. In patients
with jaundice, malabsorption of fat-soluble vitamins decreases the hepatic
production of vitamin K–dependent clotting factors and results in the pro-
longation of prothrombin time. This coagulopathy can be normalized by
the administration of parenteral or subcutaneous vitamin K. Normocytic
anemia and hypoalbuminemia may reflect the chronic nutritional sequelae
of the neoplastic disease process.
The most extensively studied serum tumor marker in PDAC is the
Lewis blood group–related mucin glycoprotein, CA 19-9. The use of CA 19-9
as a diagnostic or screening marker for PDAC is limited by several factors.
CA 19-9 is only moderately accurate (sensitivity 81%) in identifying PDAC
when using the normal cutoff value of 37 units/mL. In addition, up to 15% of
individuals do not secrete CA 19-9 because of their Lewis antigen status.
Moreover, patients with small-diameter or early tumors—who would derive
the greatest survival benefit from surgical resection—often have normal CA
19-9 levels. Lastly, elevated CA 19-9 levels can be seen in benign pancre-
atic or biliary disease, and CA 19-9 is spuriously elevated in the setting of
jaundice. Nevertheless, CA 19-9 may complement other diagnostic modali-
ties, such as computed tomography (CT), endoscopic retrograde cholan-
giopancreatography (ERCP), or endoscopic ultrasound (EUS), such that the
combined diagnostic accuracy for PDAC nearly approaches 100%. Severely
elevated preoperative CA 19-9 levels (>1,000 units/mL) suggest advanced
0002086373.INDD 41 7/14/2014 2:43:50 PM

or unresectable disease. In such instances, many authorities recommend

staging laparoscopy even when high-fidelity imaging suggests a resectable
tumor. Finally, CA 19-9 finds its best utility in postoperative surveillance.
Increasing CA 19-9 levels after surgical resection typically indicate recur-
rence or disease progression, whereas stable or declining levels postopera-
tively indicate absence of recurrence and improved overall prognosis. CEA
has little to no additive value in securing a diagnosis of PDAC.
Investigational Diagnostic Markers

The limited diagnostic utility of CA 19-9 has prompted a search for novel
tumor biomarkers to facilitate the earlier detection of PDAC. These can
be classified as (i) serum markers : CA 242, CEA, TPA, TPS, M2-pyruvate
kinase, Mic-1, IGFBP-1a, haptoglobin, serum amyloid A, etc. and (ii) tissue
markers: K-ras, p53, mucins (MUC1/2/5), microRNAs, p21, SMAD4, Bcl-2,
etc. The first four listed tissue markers have shown promise in early stud-
ies, but none are superior to CA 19-9 nor widely available for use in clinical
practice.
Diagnostic Imaging
Imaging has assumed a crucial role in the diagnosis and stratification of
PDAC patients to stage-appropriate therapy. The goals of imaging are
identification of the primary tumor, assessment of regional invasion and
vascular or lymph node involvement, evaluation of distant metastasis, and
determination of resectability. Common imaging modalities in PDAC are
the following:
(a) Transabdominal Ultrasonography (TAUS): TAUS has an operator-
dependent sensitivity of 60% to 70% in detecting PDAC, which typically
appears as a hypoechoic mass. This modality has largely been replaced
by axial imaging.
(b) Computed Tomography (CT): Multidetector-row CT (MDCT) with three-
dimensional reconstruction is the preferred noninvasive imaging modal-
ity for the diagnosis and staging of PDAC. MDCT utilizes dual-phase
imaging in the arterial and venous phases of enhancement, acquiring
sub-3-mm slices during one 20-second breath hold. PDAC typically
appears as a hypodense mass—although it can appear isodense—in the
parenchyma and is best seen on the portal venous phase of enhance-
ment. MDCT also provides a comprehensive view of tumor abutment or
encasement of the major peripancreatic vascular structures (celiac axis,
superior mesenteric artery [SMA] and superior mesenteric vein [SMV],
splenic artery, and portal vein [PV]) as well as peripancreatic lymphade-
nopathy and hepatic or omental metastasis. Finally, it provides the best
spatial road map for the anatomic relationships between the primary
tumor and the local vasculature.
(c) Magnetic Resonance Imaging (MRI): Several technical improvements
in MRI technology have improved its ability to diagnose and stage
pancreatic cancer. A majority of PDACs have low-signal intensity on
T1-weighted fat-suppressed images. Upon dynamic imaging after gado-
linium contrast injection, PDAC enhances relatively less than surround-
ing parenchyma and reveals progressive enhancement on subsequent
phases. MR cholangiopancreatography (MRCP) can optimally assess
the biliary and pancreatic ductal anatomy as well as the postresection
pancreatic remnant. Although some authorities consider MRI com-
parable to MDCT in detecting PDAC, obtaining both studies offers no
significant advantage in assessing patients with radiographically resect-
able disease. MRCP is more useful for evaluating cystic masses of the
pancreas.
002086373.INDD 42 7/14/2014 2:43:50 PM

(d) Positron Emission Tomography (PET): In this functional imaging tech-

nique, a positron-emitting tracer—fluorine-18—is labeled to fluorode-
oxyglucose (FDG), which is rapidly taken up by tumor cells. Although
unenhanced PET/CT has a limited role in the local staging of PDAC
due to its poor depiction of primary tumor and its relationship to adja-
cent vasculature, intravenous contrast-enhanced PET/CT has been
shown to have improved accuracy for PDAC detection in recent stud-
ies. However, since FDG also localizes at sites of inflammation and
infection, PET imaging is largely used as an adjunct modality to detect
disseminated disease. Furthermore, its reimbursement is currently
limited for PDAC, as opposed to its acceptance for other established
solid malignancies.
(e) Endoscopic Ultrasound (EUS): EUS has gained popularity in recent years
as a minimally invasive modality for the local staging of PDAC. Its par-
ticular benefit is realized in the clarification of small (< 2 cm) lesions in
the setting of negative or equivocal CT findings, evaluation of malignant
lymphadenopathy, detection of vascular involvement, and the ability to
obtain a tissue diagnosis when combined with fine needle aspiration
(FNA). Combined with FNA, EUS has a sensitivity of 93% and 88% for
T- and N-stage disease, respectively. However, unless protocol-based
neoadjuvant therapy is planned or the patient is a poor operative can-
didate, a tissue diagnosis is not required in acceptable-risk patients with
resectable lesions on noninvasive axial imaging (CT or MRI). Moreover,
EUS is invasive, is operator dependent, and cannot assess liver or distant
metastasis, thereby limiting its utility.
(f) Endoscopic Retrograde Cholangiopancreatography: Although its sen-
sitivity for the diagnosis of PDAC approaches 90%, the routine use of
diagnostic ERCP for suspected PDAC is unsupported. With the current
advances in CT/MRI technology, diagnostic ERCP should be reserved for
select groups of patients: (i) those with obstructive jaundice but with-
out a detectable mass on CT or MRI; (ii) symptomatic but nonjaundiced
patients with equivocal findings on CT or MRI; and (iii) suspected pan-
creatic cancer in the setting of chronic pancreatitis. Even in these circum-
stances, many authorities advocate for noninvasive alternatives, such as
MRCP. Nevertheless, ERCP is often utilized prior to surgical consultation
for relief of jaundice.
Clinicopathologic Staging
The American Joint Committee on Cancer (AJCC) staging for pancreatic
cancer is based on the TNM classification, incorporating extent of primary
tumor (T), presence of regional lymph node involvement (N), and presence
of distant metastasis (M) (Table 4.1).
MANAGEMENT
Preoperative Considerations
Assessing Surgical Resectability
Although the TNM classification is used for pathologic staging and
clinical description, the crucial step for the surgeon in the preoperative
workup of PDAC patients is to determine operability—an option available
to fewer than a quarter of all patients at diagnosis. PDAC of the pancreatic
head, neck, or uncinate process is stratified into the following: (a) resect-
able, defined as no radiographic evidence of extrapancreatic disease, a
patent SMV–PV confluence, and no evidence of tumor extension to the
celiac axis or SMA; (b) borderline resectable, defined as nonmetastatic
tumor with SMA abutment (≤180 degrees), abutment or encasement of
0002086373.INDD 43 7/14/2014 2:43:50 PM

TABLE
AJCC Staging of Pancreatic Ductal Adenocarcinoma
4.1
Tumor (T)
(seventh edition)
Tx: Primary tumor cannot be assessed

T0: No evidence of primary tumor
Tis: Carcinoma in situ
T1: Tumor limited to the pancreas, 2 cm or less in greatest dimension
T2: Tumor limited to the pancreas, more than 2 cm in greatest dimension
T3: Tumor extends beyond the pancreas but without involvement of the celiac
axis or the superior mesenteric artery
T4: Tumor involves the celiac axis or the superior mesenteric artery
(unresectable primary tumor)
Regional Lymph Nodes (N)
Nx: Regional lymph nodes cannot be assessed
N0: No regional lymph node metastasis
N1: Regional lymph node metastasis
Distant Metastasis (M)
Mx: Distant metastasis cannot be assessed
M0: No distant metastasis
M1: Distant metastasis
Stage T N M 5-Year Survival (%)
IA T1 N0 M0 20–30
IB T2 N0 M0 20–30
IIA T3 N0 M0 10–25
IIB T1, T2, T3 N1 M0 10–15
III T4 Any N M0 0–5
IV Any T Any N M1 —
the gastroduodenal artery (GDA) up to or around a short, reconstructable

segment of the common hepatic artery (CHA), or complete occlusion of
the SMV–PV confluence, with suitable, uninvolved vein above and below
to allow venous reconstruction (as a general rule, tumors encasing up to
2 cm of the SMV and/or PV for up to a 180-degree circumference are con-
sidered technically possible); (c) locally advanced, defined as encasement
(>180 degrees) of the celiac axis or SMA or occlusion of SMV–PV conflu-
ence without an option for venous resection and reconstruction; and
(d) metastatic, defined as distant metastatic spread to the liver, perito-
neum, or rarely the lung. It is important to realize that, for this disease,
operability ≠ resectability. Up to 85% (but certainly not 100%) of radiograph-
ically amenable lesions are resectable intraoperatively, whereas fewer than
20% of locally advanced lesions are ultimately amenable to resection.
Role of Diagnostic Laparoscopy

The use of diagnostic laparoscopy in PDAC staging remains controversial.
Although the combination of laparoscopy with laparoscopic ultrasound can
improve the accuracy of determining peritoneal and/or hepatic dissemina-
tion to almost 98%, we do not recommend the routine application of diagnos-
tic laparoscopy in PDAC due to its questionable cost-effectiveness as well as
the impressive accuracy of modern high-fidelity axial imaging. Moreover, even
in potentially unresectable disease, laparoscopy is unnecessary if the patient
were to benefit most from surgical palliation (see below). Alternatively, a
more selective use of diagnostic laparoscopy is justified in those patients with
a high risk of occult metastatic disease in whom nonsurgical palliation would
002086373.INDD 44 7/14/2014 2:43:50 PM

be durable. This scenario includes the following: (a) lesions in the pancreatic
body or tail; (b) larger (> 3 cm) primary tumors; (c) equivocal radiographic
findings suggestive of occult metastasis such as low-volume ascites, perito-
neal carcinomatosis, or small hepatic lesions not amenable to percutaneous
biopsy; and (d) markedly elevated CA 19-9 or hypoalbuminemia suggestive
of advanced disease. In addition, diagnostic laparoscopy may be considered
in certain cases of borderline resectable and/or locally advanced disease—
understanding that evidence of metastatic spread might alter whether radia-
tion therapy would be used in conjunction with chemotherapy.
Role of Preoperative Biliary Drainage for Obstructive Jaundice

In the past, a majority of patients with resectable PDAC who presented
with obstructive jaundice underwent either endoscopic or percutaneous
transhepatic preoperative biliary drainage (PBD) before surgical referral.
Recently, however, the routine use of PBD in resectable PDAC has been dis-
couraged in the light of convincing evidence that PBD increases the rate of
postoperative complications and offers little benefit compared with early
surgery for resectable tumors. In a recent randomized multicenter trial, van
der Gaag et al. showed that PBD was associated with significantly higher
rates of serious perioperative complications compared with early surgery,
while mortality and length of hospital stay did not differ significantly
between the groups. However, the role of biliary drainage as temporary
palliation in patients with borderline resectable tumors undergoing neo-
adjuvant therapy is clinically relevant. In such situations, authorities advo-
cate the use of self-expandable metal stents over plastic stents given their
improved durability should the tumor be ultimately deemed unresectable.
Role of Neoadjuvant Therapy for PDAC

There is growing enthusiasm for the use of neoadjuvant chemoradiation
in PDAC that relies on the following theoretical principles: (a) increased
efficacy of chemoradiation in well-oxygenated tissue not devascularized by
surgery; (b) potential for downstaging, thereby improving R0 resection and
decreasing locoregional recurrence rates; and (c) identifying patients with
rapidly progressive disease who might not derive a survival advantage from
surgery. Several groups have utilized neoadjuvant regimens involving gem-
citabine, 5-FU, or paclitaxel plus radiation for resectable PDAC. Current
data suggest that while neoadjuvant chemotherapy is well tolerated and
does not delay surgical intervention, there is no clear survival benefit to this
strategy compared with standard postoperative adjuvant therapy. A 2009
AHPBA-SSO-SSAT consensus panel indicated that neoadjuvant therapy for
resectable PDAC should be considered investigational and offered within
the context of clinical trials or multidisciplinary treatment programs.
However, there may be an emerging role for neoadjuvant strategies in
borderline resectable tumors. Investigators from the University of Virginia
demonstrated that 46% of patients with borderline resectable PDAC under-
went surgical resection after completing neoadjuvant capecitabine-based
chemoradiation. This group had favorable overall survival (OS)—median
23 months—similar to that of a historical cohort of patients with resectable
PDAC. In a recent report from the M.D. Anderson Cancer Center, 66% of
borderline resectable PDAC patients underwent resection after neoadju-
vant therapy (chemoradiation alone or gemcitabine-based chemotherapy
followed by chemoradiation), of which an R0 resection was achieved in
95%. Median OS in this cohort was up to 33 months following their uniquely
highly standardized protocol. These data solidify the role for induction
therapy in borderline resectable PDAC, with attempted resection offered to
patients without evidence of cancer progression on presurgical restaging
0002086373.INDD 45 7/14/2014 2:43:50 PM

studies. Several challenges remain: (a) adherence to and application of the

consensus definition of “borderline resectable” tumors across institutions;
(b) what the most effective neoadjuvant modality is (chemotherapy alone,
chemoradiation, or a combination); and (c) need for a multi-institutional
trial to validate these optimistic early reports. As is the case for resectable
tumors, it remains a fact that there has been no direct clinical comparison
of the two techniques that would clearly advocate one over the other.
Surgical Resection and Related Controversies

Resection of PDAC can be broadly categorized into two types: (i) pancreati-
coduodenectomy for tumors situated in the head, neck, or uncinate process
or (ii) distal pancreatectomy for tumors of the body and tail. While techni-
cal details of these operations can be found in Chapter 7 by Pilgrim et al.,
a discussion of the global conduct of these operations and select lingering
controversies in the operative management of PDAC follows.
Pancreaticoduodenectomy (Whipple Procedure)

The preliminary steps of a PD or Whipple procedure involve assessment of
resectability. The liver, peritoneal surfaces, celiac axis–level lymph nodes,
and remainder of the abdomen are inspected for metastatic disease. As
noted earlier, this assessment may be performed laparoscopically in select
circumstances. A Kocher maneuver is performed to assess for retroperito-
neal, SMV, SMA, or celiac takeoff involvement. The SMV is identified in the
plane anterior to the third portion of the duodenum between the transverse
mesocolon and the uncinate process by opening the lesser sac through the
gastrocolic omentum. The porta hepatis is examined by mobilizing the gall-
bladder and tracing the cystic duct to its junction with the common hepatic
duct. Once the tumor is considered resectable on both of these fronts, early
division of the extrahepatic biliary tree allows caudal retraction of the distal
common bile duct and visualization of the anterior aspect of the PV. In a
pylorus-preserving PD, division of bowel occurs 2 cm distal to the pylorus
and 20 cm distal to the ligament of Treitz. The divided proximal jejunum is
delivered posterior to the superior mesenteric vessels to facilitate dissec-
tion of the pancreas off the right lateral aspect of the SMV. Vein resection
may be required and ranges from primary repair to patch reconstruction
to interposition grafting. The pancreatic neck overlying the SMV–PV is
divided, and the pancreatic head is meticulously dissected off the right
lateral aspects of the SMV–PV. Dissection of the retroperitoneal “meso-
pancreas” ensues with the aim of removing the tissue up to the adventitial
layer of the SMA along the right side of its course. Reconstruction involves
a pancreaticoenteric anastomosis—most commonly pancreaticojejunos-
tomy (PJ)—in an end-to-end or end-to-side fashion using either a duct-to-
mucosa or invagination technique. Alternatively, a pancreaticogastrostomy
(PG) can be used. The biliary–enteric anastomosis is typically performed in
an end-to-side fashion approximately 10 cm distal to the PJ on the pancre-
aticobiliary limb of the jejunum. The duodenojejunostomy (DJ) is placed, in
a majority of cases, in an antecolic orientation 50 to 60 cm downstream to
the biliary–enteric anastomosis. Enteric feeding tubes (gastrostomy or jeju-
nostomy) can be used selectively in those patients who the surgeon expects
may not thrive postoperatively.
Fewer resections are performed for tumors of the body and tail of the pan-
creas due to a higher incidence of advanced disease at initial presenta-
tion. Resection of these tumors is achieved with a distal pancreatectomy
accompanied by, in a vast majority of cases, splenectomy. As with PD, initial
002086373.INDD 46 7/14/2014 2:43:50 PM

exploration is geared to rule out hepatic or peritoneal metastasis, and lapa-

roscopy has a higher yield than with proximal tumors. If metastasis is not
present, the lesser sac is entered and involvement of the SMV is adjudicated.
If resection is attempted, the spleen and distal pancreas are mobilized in a
retrograde fashion. The pancreas is then transected to the left of the SMV–
PV trunk and the main pancreatic duct ligated. This is usually performed
with a stapling device, although suturing the duct is acceptable—evidence
does not support any particular technique. However, due to limitations in
the posterior extent of resection and the ability to achieve a complete N1
nodal dissection with the classic distal pancreatosplenectomy, Strasberg
et al. proposed the radical antegrade modular pancreatosplenectomy
(RAMPS) procedure in 2003. This approach, which relies on medial to lateral
dissection, improves visibility, enables an N1 nodal dissection, enhances the
ability to achieve microscopically negative tangential margins, and permits
adjustment of the depth of the posterior extent of resection coupled with
early rather than late control of the vasculature. Long-term results revealed
R0 resection rates of up to 81%, negative tangential margins in 89%, and
median survival of 26 months.
Surgical Palliation
PDAC patients found to have unresectable disease at laparotomy, or those
with tumor-related symptoms poorly alleviated by nonoperative methods,
are appropriate candidates for surgical palliation. It involves biliary bypass,
gastroenteric bypass, chemical splanchnicectomy, or combinations thereof.
Hepatico- or choledochojejunostomy, using a Roux-en-Y conduit, is the pre-
ferred method of biliary bypass and results in a decreased risk of recurrent
jaundice compared with choledochoduodenostomy or cholecystojejunos-
tomy. Some patients undergoing biliary bypass alone ultimately require
gastroenteric bypass for eventual duodenal obstruction. A prospective trial
evaluating the role of prophylactic gastrojejunostomy (GJ) in unresectable
PDAC found that 19% of patients randomized to no GJ developed late gas-
tric outlet obstruction requiring intervention compared with 0% in the GJ
group (p < 0.01), although mean OS between the groups was similar. Finally,
chemical splanchnicectomy—injection of 50% alcohol at the celiac nerve
plexus—significantly reduces mean pain scores in patients with unresect-
able right-sided PDAC. This should be reserved for unresectable patients
who present with significant pain prior to laparotomy.
Controversies in Pancreatic Resection for PDAC

(a) Role for Extended Lymphadenectomy: Four randomized trials have
attempted to answer if there is benefit to extended lymph node dissec-
tion in PDAC. The data, substantiated by numerous strong meta-analyses,
suggest that extended lymphadenectomy increases the morbidity without
significantly improving median or 5-year survival.
(b) Classical Versus Pylorus-Preserving PD: Most surgeons prefer the pylorus-
preserving PD because it reduces operative time and blood loss, retains
the entire gastric reservoir, and does not compromise oncologic out-
comes (negative margin rate or OS) compared with the classical variant.
Furthermore, no appreciable difference is observed in gastric physiology
or nutrition postoperatively.
(c) Vascular Resection and Reconstruction: The data examining vascular
resection and reconstruction after PD for PDAC are nonstandardized.
Several single-center experiences suggest that, in the absence of other
contraindications for resection, a controlled venous resection and
reconstruction when the tumor cannot be separated from the SMV, PV
or SMV–PV confluence is safe and feasible. Recently, however, a large
0002086373.INDD 47 7/14/2014 2:43:50 PM

retrospective analysis of the ACS NSQIP database revealed that vascu-

lar resection after PD is associated with a significantly higher overall
morbidity and 30-day postoperative mortality. Arterial resection is not
advised with the exception of a short segment of hepatic artery for those
tumors invading the GDA takeoff. Most experienced pancreatic surgeons
avoid concomitant arterial and venous resection.
(d) Margin Status: Current margin nomenclature is poorly defined and the
value of intraoperative frozen section analysis has not been studied
extensively. Controversy exists over the proper techniques for speci-
men analysis by pathologists. Achieving an R0 resection and a negative
SMA/uncinate margin correlates most strongly with positive long-term
outcomes. Patients with R1 resections should be strongly considered for
adjuvant multimodal therapy.
(e) Minimally Invasive PD: Although minimally invasive distal pancreatic
resections are being increasingly employed for cancer, minimally invasive
PD (MIPD) for PDAC continues to lack widespread acceptance because
matched comparative trials and long-term outcomes comparing MIPD
to the open approach are lacking. Recently, comparisons between the two
suggest that MIPD is associated with reduced length of stay and blood
loss without compromising standard oncologic principles. Although
long-term follow-up is lacking, short-term results suggest equivalent, but
not better, rates of recurrence and survival when highly skilled specialists
apply MIPD techniques.
Postoperative Considerations
Postoperative Complications
The overall complication rate after PDAC resection, especially PD, con-
tinues to approach 40%. The most common immediate complications are
postoperative pancreatic fistula (approximately 15%), delayed gastric emp-
tying (approximately 5%), hemorrhage (approximately 5%), and wound
infection (approximately 20%). Long-term effects of parenchymal resection
may manifest in pancreatic insufficiency. These include diabetes and steat-
orrhea—both of which occur in approximately 25% of all patients.
(a) Postoperative Pancreatic Fistula (POPF): A gradation of POPF—A (bio-
chemical, with no clinical impact) through C (rendering a life-threatening,
major deviation from clinical recovery)—was developed by a consensus
panel (ISGPF) in 2005. There is no unequivocal evidence to suggest that
anastomotic route (PJ vs. PG), anastomotic technique (duct-to-mucosa
vs. invagination vs. single-layer end-to-side PJ), pancreatic duct stenting,
or use of long-acting somatostatin analogs alters the incidence of POPF
formation.
(b) Delayed Gastric Emptying (DGE): The severity of DGE after pylorus-
preserving PD—grades A through C—relates to the need for nasogas-
tric decompression, volume of gastric effluent, and inability to tolerate
oral intake postoperatively. The pathogenesis of DGE remains unclear,
although putative mechanisms include antral ischemia, absence of duo-
denal motilin, gastric atony from a vagotomized stomach, and gastric
dysrhythmia secondary to POPF or edema from fluid overload. There
is convincing randomized evidence that (i) an antecolic compared
with a retrocolic reconstruction and (ii) postoperative administration
of erythromycin (motilin agonist) significantly reduce the incidence of
DGE after pylorus-preserving PD.
Postoperative Adjuvant Therapy
Adjuvant therapy, supported by a large body of evidence, is a critical part of
the postoperative management of PDAC. The Gastrointestinal Tumor Study
002086373.INDD 48 7/14/2014 2:43:50 PM

Group (GISTG, 1985) originally showed that 5-FU–based chemoradiation

followed by maintenance chemotherapy had a significant median OS ben-
efit compared with observation alone (20 months vs. 11 months, p = 0.03).
Despite limited accrual, the GITSG trial was the first to show a potential ben-
efit for adjuvant therapy after resection of pancreatic cancer. The European
Organization for Research and Treatment of Cancer (EORTC, 1999) study
was similar and randomized resected patients to concurrent chemoradia-
tion without subsequent chemotherapy versus observation alone. Median
OS was not significantly different (24.5 months vs. 19 months, p = 0.208).
The European Study Group for Pancreatic Cancer (ESPAC-1, 2004) study
was conducted with a 2 × 2 factorial design. The four arms included sur-
gery alone, 5-FU–based chemotherapy alone, 5-FU–based chemoradiation
alone, and both modalities. Patients receiving chemotherapy (chemother-
apy alone or chemoradiation followed by chemotherapy) had a significantly
improved median OS compared with the no-chemotherapy arm (chemo-
radiation alone or surgery alone; 19.7 months vs. 14 months, p = 0.0005).
Interestingly, patients who received radiation (chemoradiation alone or
chemoradiation followed by chemotherapy) had worse median OS com-
pared with those not receiving radiation (chemotherapy alone and surgery
alone; 15.9 months vs. 17.9 months, p = 0.05).
With gemcitabine emerging as a viable alternative to 5-FU chemo-
therapy, the CONKO-001 study demonstrated that resected patients receiv-
ing adjuvant gemcitabine had significantly improved median OS compared
with observation alone (22.8 months vs. 20.2 months, p = 0.005). The ESPAC-3
trial, a head-to-head comparison between adjuvant gemcitabine versus 5-FU
alone, found a nonsignificant difference in median OS (23.6 months vs. 23.0
months). Of note, there was a greater incidence of high-grade, treatment-
related toxicities in the 5-FU arm. The Radiation Therapy Oncology Group
(RTOG-9704), which compared adjuvant gemcitabine + 5-FU–based chemo-
radiation and 5-FU + 5-FU–based chemoradiation, found a nonstatistically
significant survival benefit with gemcitabine (20.5 months vs. 16.9 months,
p = 0.09). Although both 5-FU and gemcitabine have been validated as via-
ble adjuvant options, gemcitabine has largely replaced 5-FU in the Unites
States due its greater tolerability and improved quality-of-life outcomes.
The Debate Surrounding Adjuvant Radiotherapy
Although the benefit of adjuvant chemotherapy is unequivocally agreed
upon, the concurrent use of radiation therapy remains contentious. In the
United States, the use of adjuvant chemoradiation is still largely standard
practice. Proponents for this approach cite the survival benefit seen in the
GISTG trial and the design flaws underlying the conclusions of the ESPAC-1
trial. In Europe, adjuvant chemotherapy alone is preferred, presumably
from different interpretations of the same data presented above.
Future Directions in Adjuvant Therapy

Adjuvant trials with erlotinib, FOLFIRINOX, maintenance capecitabine, and
a therapeutic vaccine GI-4000 are underway and are described elsewhere.
OUTCOMES AND FOLLOW-UP

Historically, after PDAC resection, the perioperative mortality rate was 20%
to 40% and 5-year survival approximately 3%. These statistics have changed
dramatically, with current operative mortality rates having declined to well
under 5% and 5-year survival regularly eclipsing 20% in recent studies from
high-volume pancreatic surgery centers. This improvement is likely multi-
factorial and is related to advances in preoperative imaging, improvements
in surgical technique and perioperative care, and availability of broader
0002086373.INDD 49 7/14/2014 2:43:51 PM

adjuvant therapy options. Recently, a contemporary analysis of 424 PDAC

resections from the University of Pennsylvania and Beth Israel Deaconess
Medical Center demonstrated median, 1-year, and 5-year survivals of
21.3 months, 76%, and 23%, respectively. The 30-day mortality was only
0.7%. Seventy-six percent of patients received adjuvant therapy. Patients
with major complications (Clavien grade IIIb–IV) survived similarly to
those without complications. In general, predictors of survival include
patients with R0/N0/M0 characteristics, tumor diameters less than 3 cm,
negative resection margins, and well or moderate tumor differentiation and
those who receive adjuvant chemoradiation.
CONCLUSION
Overall, the prognosis of PDAC continues to be disappointing. However,
survival in resectable PDAC has increased dramatically over the last few
decades, in part due to significant improvements in surgical expertise and
adjuvant multimodal therapy. Correspondingly, the therapeutic boundaries
in PDAC care are being pushed with respect to expanding indications for
PDAC resection and aggressive neoadjuvant strategies in borderline resect-
able tumors. Continuing refinement in our surgical techniques, combined
with emerging chemo- and radiotherapeutic options, provides cautious
optimism for the future of this often frustrating disease.
Suggested Readings
Bardeesy N, DePinho RA. Pancreatic cancer biology and genetics. Nat Rev Cancer
2002;2(12):897–909.
Castleberry AW, White RR, De La Fuente SG, et al. The impact of vascular resection on
early postoperative outcomes after pancreaticoduodenectomy: an analysis of the
American College of Surgeons National Surgical Quality Improvement Program
database. Ann Surg Oncol 2012;19(13):4068–4077.
Diener MK, Knaebel HP, Heukaufer C, et al. A systematic review and meta-analysis of
pylorus-preserving versus classical pancreaticoduodenectomy for surgical treat-
ment of periampullary and pancreatic carcinoma. Ann Surg 2007;245(2):187–200.
Hruban RH, Brune K, Klein A, et al. Familial pancreatic cancer. In: Von Hoff DD, Evans
DB, Hruban RH (eds.), Pancreatic cancer. Boston, MA: Jones and Bartlett, 2004.
Katz MH, Fleming JB, Bhosale P, et al. Response of borderline resectable pancreatic
cancer to neoadjuvant therapy is not reflected by radiographic indicators. Cancer
2012;118(23):5749–5756.
Kawai M, Yamaue H. Analysis of clinical trials evaluating complications after pan-
creaticoduodenectomy: a new era of pancreatic surgery. Surg Today 2010;40(11):
1011–1017.
Kendrick ML. Laparoscopic and robotic resection for pancreatic cancer. Cancer J
2012;18(6):571–576.
Lewis R, Drebin JA, Callery MP, et al. A contemporary analysis of survival for resected
pancreatic ductal adenocarcinoma. HPB (Oxford) 2013;15(1):49–60.
Mitchem JB, Hamilton N, Gao F, et al. Long-term results of resection of adenocarci-
noma of the body and tail of the pancreas using radical antegrade modular pan-
creatosplenectomy procedure. J Am Coll Surg 2012;214(1):46–52.
Smaglo BG, Pishvaian MJ. Postresection chemotherapy for pancreatic cancer. Cancer J
2012;18(6):614–623
Stokes JB, Nolan NJ, Stelow EB, et al. Preoperative capecitabine and concurrent radia-
tion for borderline resectable pancreatic cancer. Ann Surg Oncol 2011;18:619–627.
Van der Gaag NA, Rauws EA, van Eijck CH, et al. Preoperative biliary drainage for
cancer of the head of the pancreas. N Engl J Med 2010;362:129–137.
Zamboni GA, Kruskal JB, Vollmer CM, et al. Pancreatic adenocarcinoma: value of
multidetector CT angiography in preoperative evaluation. Radiology 2007;245(3):
770–778.
002086373.INDD 50 7/14/2014 2:43:51 PM

5 Pancreatic Cysts Including
Intraductal Pancreatic
Mucinous Neoplasm (IPMN)
Joshua A. Waters and C. Max Schmidt
ETIOLOGY
Few entities within the practice of pancreatic surgery have garnered more
interest and discussion in the last decade than pancreatic cystic lesions.
Cystic lesions of the pancreas can be divided broadly into inflammatory
and neoplastic categories. Inflammatory cysts of the pancreas, primarily
comprised of pseudocysts, will be discussed elsewhere in this handbook.
The objective of this chapter is to characterize pancreatic cystic lesions that
arise of neoplastic etiology.
Cystic lesions of the pancreas, though once believed to be rare, may
be present in up to 25% of patients based on autopsy studies and up to 15%
of patients based on cross-sectional imaging studies. Although there are
numerous ways to subdivide pancreatic cystic lesions, the most clinically
relevant way to further categorize cystic pancreatic lesions is based on their
malignant potential. Three broad categories commonly used to describe
pancreatic cysts lesions are benign, premalignant, or malignant.
Benign
Although benign from an oncologic standpoint, benign cysts may grow to
cause local compressive symptoms or other complications of mass effect.
They are also important to recognize because they may mimic premalig-
nant or malignant lesions of the pancreas.
Serous cystic neoplasms (SCNs), also known historically as serous cystad-
enomas (SCA), are the most common benign cystic neoplasm of the pancreas.
These lesions are lined by glycogen-rich cuboidal epithelium. These lesions
are most common in women (70%) and usually occur sporadically. In some
patients, they can be multifocal, but this multifocality is most common in
association with the autosomal dominant von Hippel-Lindau syndrome. SCN
may be macrocystic (Fig. 5.1) or microcystic with multiple small septations.
Microcystic SCN often have a characteristic honeycomb appearance allowing
them to be more readily recognized on cross-sectional and endoscopic ultra-
sound (EUS) imaging. Importantly, however, tightly packed multifocal branch
duct intraductal papillary mucinous neoplasm (IPMN) can mimic microcys-
tic SCN, making it sometimes indistinguishable. Although less common, the
macrocystic variety of SCN can notoriously mimic mucinous cystic lesions
(both mucinous cystic neoplasms [MCN] and IPMN). Lymphoepithelial cysts
are much rarer than their benign counterparts and more commonly occur in
men. These lesions are lined with keratinized squamous epithelium and have
a characteristic exophytic appearance on cross-sectional imaging (Fig. 5.2).
They have been known to erode into surrounding structures, which often ren-
ders these cysts symptomatic, and they may sometimes appear malignant in
nature. Finally, simple pancreatic cysts are unilocular and invariably benign.
These cysts may have some association with polycystic disease of other
organs (kidney, liver), but do not demonstrate a gender or age predilection.
51
0002086374.INDD 51 7/14/2014 2:51:48 PM

FIGURE 5.1 CT scan demonstrating a macrocystic serous cystic neoplasm (SCN) within
the body of the pancreas.
Premalignant and Malignant

The clinical importance of identification of premalignant/malignant pan-
creatic cysts cannot be overstated. Their identification represents an
opportunity to intervene prior to the development of invasive pancreatic
malignancy. That said, the majority of premalignant pancreatic cysts will
FIGURE 5.2 Arterial-phase CT scan demonstrating the characteristic exophytic appearance

of a pancreatic lymphoepithelial cyst.
002086374.INDD 52 7/14/2014 2:51:49 PM

Chapter 5 / Pancreatic Cysts Including Intraductal Pancreatic Mucinous Neoplasm (IPMN) 53
not transform into pancreatic cancer in a given patient’s lifetime, so

surveillance is often recommended. Pancreatic cysts in the premalignant/
malignant category include mucinous cysts (MCN and IPMN), cystic neuro-
endocrine tumors, and solid pseudopapillary neoplasm (SPN).
Mucinous cystic neoplasms (MCNs) occur primarily in female patients
in the fifth to sixth decade of life. These are mucin-filled lesions without
any connection to the pancreatic ductal system. They are unifocal and
macrocytic with histology demonstrating ovarian-type stroma. They may
demonstrate calcifications. When MCN transform, they become mucinous
cystadenocarcinoma, which may represent up to 10% to 20% of MCN. Early
in their course, mucinous cystadenocarcinoma may represent a more indo-
lent form of cancer when compared to pancreatic adenocarcinoma. These
lesions may stain positive for human chronic gonadotropic or estrogen
receptors. Intraductal papillary mucinous neoplasm (IPMN) represents a
spectrum of lesions with several common features including pancreatic
ductal connection, mucin-producing epithelium, and papillary histology
(Fig. 5.3). These cystic lesions may be unifocal, multifocal, or multicentric
and occur most commonly in older patients (sixth to seventh decade).
Anatomically, IPMN may involve the main or branch pancreatic ducts or
both. Invasive IPMN is the malignant endpoint in the IPMN oncologic spec-
trum. Malignant lesions may represent around 25% of all resected IPMN
specimens and fall into two distinct histologic subtypes. The tubular sub-
type imitates the appearance and behavior of pancreatic ductal adeno-
carcinoma, whereas the colloid subtype is characterized by large pools of
extracellular mucin and follows a relatively more indolent course. Cystic
pancreatic neuroendocrine tumors (pNETs) are rare variants of typical
pNET and may occur anywhere on the spectrum from benign to malignant
behavior. These lesions occur more commonly in women and are typically
sporadic, though they may exist as part of a multiple endocrine neoplasia
syndrome. Solid and cystic pseudopapillary neoplasms (SPNs) are rare het-
erogenous solid and cystic lesions that occur almost exclusively in younger
FIGURE 5.3 Hematoxylin and eosin stain histologic operative specimen demonstrating the
prominent mucin-laden papillary epithelial architecture consistent with IPMN.
0002086374.INDD 53 7/14/2014 2:51:51 PM

women. These lesions represent a spectrum from premalignant to

malignant (about 10%), but generally have an indolent course.
DIAGNOSIS
The explosion in interest surrounding cystic lesions of the pancreas is due
in part to their increased recognition through improved resolution and
broader use of contemporary cross-sectional imaging. In addition, there is
an increased awareness of the importance of pancreatic cyst identification
and characterization to promote early detection, prevention, and potential
cure of pancreatic cancer. A comprehensive approach to evaluation and
characterization of these lesions is critical to determine: first, the type of
cystic lesion and, second, the relative oncologic risk of the lesion. Diagnosis
relies on clinical, serologic, radiologic, cytopathologic, and cyst fluid (bio-
chemical and DNA) evaluation.
Clinical
Obtaining a detailed clinical history and physical examination plays an
important role in the diagnosis and differentiation of pancreatic cystic
lesions. As highlighted above, the epidemiologic patterns of these cysts with
regard to age and sex may aid in the development of the initial differential
diagnosis (e.g., MCN occurring primarily in females). Symptoms attribut-
able to any of these lesions may be subtle. In observational studies, the most
commonly reported symptoms include epigastric pain or fullness, early
satiety, and sometimes nausea. Evaluation for symptoms related to pancre-
atic exocrine and endocrine dysfunction is also important. Compression or
occlusion of the pancreatic ductal system secondary to a cystic lesion may
result in a spectrum of exocrine pancreatic dysfunction and ultimately fail-
ure. Symptoms of exocrine insufficiency may initially present as bloating
and flatulence, but ultimately may progress to steatorrhea, malabsorption,
and weight loss. Pancreatic duct obstruction may also precipitate pancre-
atitis. Derangements in insulin production may lead to new-onset or wors-
ening diabetes mellitus in some patients. As with most pancreatobiliary
malignancies, the presence of jaundice and cachexia is an ominous sign,
which may indicate advanced malignancy. Physical examination will typi-
cally be somewhat low yield in these lesions. These patients rarely present
with reproducible epigastric tenderness on exam, but a palpable abdominal
mass may be present in the setting of larger lesions.
Serologic
The use of laboratory studies in the diagnosis and workup for cystic lesions
of the pancreas may be of some adjunctive value. Assay for the develop-
ment of new or worsening diabetes (serum hemoglobin A1c, fasting serum
glucose) as well as pancreatic enzymes (serum amylase and lipase) to assess
for pancreatitis in the setting of epigastric pain and nausea should be per-
formed. Additionally, serum tumor markers for pancreaticobiliary malig-
nancy, particularly cancer antigen (CA) 19-9, should be checked, as this has
been shown to correlate with presence of invasive disease in the setting of
IPMN. Elevation of serum alkaline phosphatase has been correlated with
malignancy in pancreatic head cysts well prior to elevation of bilirubin. As
of yet, there are no other viable serologic biomarkers for diagnosis or deter-
mination of malignant potential of pancreatic cystic lesions.
Radiologic
Cross-sectional imaging is the workhorse for the initial diagnosis and
characterization of pancreatic cysts. Although debate exists as to their
002086374.INDD 54 7/14/2014 2:51:51 PM

relative value, computed tomography (CT) and magnetic resonance

imaging/magnetic resonance cholangiopancreatography (MRI/MRCP) are
the two primary noninvasive imaging modalities utilized for characteriza-
tion these lesions.
Regardless of the modality utilized, certain radiologic characteristics
of pancreatic cystic lesions may suggest a particular etiology. Two elements,
with rare exception, that are predictive of IPMN are ductal connection and
multifocality. These two features are extremely rare in other cystic lesions.
A central stellate scar is a classic finding for SCN, and mural calcifications
may often be present in (but are not pathognomonic of) MCN (Fig. 5.4).
Finally, in the setting of malignant degeneration, cystic lesions of the pan-
creas on cross-sectional imaging may demonstrate features of invasion into
extrapancreatic organs, vascular structures, or the biliary tree. Peritoneal
and hematogenous metastases may also be detected in this setting.
Dual-phase intravenous contrast-enhanced computed tomography
(CT) scan is the mainstay for initial diagnosis and characterization of pan-
creatic cysts. Many pancreatic cysts (up to 30%) will actually be identified as
incidentally detected lesions on CT scan performed for other indications.
Current high-resolution CT scan technology allows for very accurate char-
acterization of these lesions. Most cystic lesions will appear hypodense with
respect to the surrounding pancreatic parenchyma on CT. It is critical to
obtain thin (1 mm) cuts through the pancreas as smaller lesions are easily
overlooked with less detailed imaging.
Magnetic resonance cholangiopancreatography (MRCP) is another non-
invasive imaging modality that has gained wide acceptance as important
adjunct in characterizing pancreatic cysts. The ability to obtain a detailed
pancreatic ductogram (Fig. 5.5) may provide the clinician with information
regarding ductal connection (MRI is two to three times more sensitive than
CT scan alone) and may be more accurate in delineating subtle main pan-
creatic duct (MPD) involvement. Some authors suggest that MRCP is supe-
rior in defining IPMN type and extent and routinely employ this technique
in preoperative planning and surveillance strategies.
FIGURE 5.4 MCN of the pancreas with the associated mural calcification seen on CT.
0002086374.INDD 55 7/14/2014 2:51:51 PM

FIGURE 5.5 MRCP illustrating the grapelike morphology of a branch-type IPMN with an
obvious ductal connection.
Positron emission tomography (PET) scan has been examined as an

adjunctive imaging study to determine the presence of 18fludeoxyglucose
(FDG) avid foci within IPMN particularly as a marker for invasive malig-
nancy. Currently, PET scanning is used to help with clinical decision making
in cases where there is a high index of suspicion for malignancy concomi-
tant with a moderate to high risk of surgical intervention. The sensitivity of
PET scan for high-grade dysplastic lesions and small foci of invasive cancer
is limited, resulting in a significant false-negative rate.
Although noninvasive imaging techniques are the frontline in the
diagnosis of pancreatic cystic lesions, endoscopic ultrasound (EUS) has
become very important in the detailed evaluation for many pancreatic
cysts. Although EUS is highly operator dependent, this technique allows for
detailed characterization of cyst size, number, relationship with vascular
and ductal structures, and presence of an associated mass or mural nod-
ule. The presence of mural nodule or associated mass (Fig. 5.6) in the set-
ting of IPMN in particular has been reproducibly demonstrated as a strong
predictor of underlying malignancy, and EUS is the most accurate method
of identifying and characterizing these features. Additionally, this modal-
ity provides the surgeon or gastroenterologist with the ability to perform
fine needle or core biopsy with assay of cyst wall or fluid. Limitations of
EUS include difficulty in fully assessing the pancreatic uncinate process, the
invasive nature of the test, and the aforementioned operator dependence.
In addition, this modality is not generally available outside of the larger
centers.
Endoscopic retrograde cholangiopancreatography (ERCP) was at one
time commonly used in the diagnosis of IPMN. It has very little role in the
workup of other cystic pancreatic lesions, as it can only evaluate cysts that
are contiguous with the MPD. Interestingly, ERCP was noted to have a
false-negative rate in the evaluation of IPMN, particularly in branch-type
002086374.INDD 56 7/14/2014 2:51:52 PM

FIGURE 5.6 EUS illustrating mural nodularity within a branch-type IPMN. This feature
alone is highly suggestive of malignancy.
lesions, where mucin plugging may preclude a full and accurate ductogram.
Additionally, there is limited ability to sample non–main duct–involved
IPMN. For these reasons, ERCP has been largely supplanted by EUS and
cross-sectional imaging techniques, especially MRI/MRCP.
Cytopathologic
Cytopathology, typically derived from fine needle aspiration (FNA)
obtained via EUS guidance, has very high specificity (95% to 100%) but
limited sensitivity (50% to 60%) for underlying malignancy in the setting of
mucinous cystic lesions (MCN and IPMN). Thus, a cytopathologic finding of
high-grade atypia is almost always indicative of an underlying malignant
lesion. Cytopathology is highly accurate (nearly 100%) in the characteriza-
tion of cystic pNET and nearly as accurate for identifying SPN. Limitations
of EUS–FNA cytopathology include difficulty in accessing the pancreatic
uncinate and the need for transgastric passage of the biopsy needle, which
may result in false-positive assays for mucin.
Cyst Fluid Analysis

One of the best ways to differentiate pancreatic cysts is via cyst fluid anal-
ysis. The most useful biochemical marker is carcinoembryonic antigen
(CEA). A high CEA level at a threshold of greater than 192 ng/mL is 80%
accurate in differentiating mucinous (IPMN, MCN) from nonmucinous
(SCN) lesions. The higher the CEA, the more certain the lesion is to be muci-
nous. This is important, as nonmucinous lesions less commonly represent
malignant risk. Absolute CEA level is not an accurate indicator of malig-
nancy. Cyst epithelial DNA shed into the cyst fluid allows for genetic char-
acterization of these lesions. A commercial test (PathFinder TG) assesses
the quantity and quality of DNA within the specimen as well as the presence
0002086374.INDD 57 7/14/2014 2:51:53 PM

TABLE
DNA Markers in Cyst Fluid and/or Cyst Epithelium According
5.1 to Pancreatic Cyst Type
Type KRAS GNAS RNF43 CTNNB1 VHL

IPMN + + +
MCN + +
SPN +
SCN +
KRAS, Kirsten rat sarcoma oncogene; GNAS, guanine nucleotide–binding protein alpha-stimulating,
complex locus; RNF43, ring finger protein 43; CTNNB1, beta-catenin gene; VHL, von Hippel-Lindau.
and copy number of Kirsten rat sarcoma viral oncogene homolog (Kras)
mutations, guanine nucleotide protein GSα (GNAS) mutations, and mic-
rosatellite instability to predict loss of heterozygosity at specific loci (e.g.,
17q). This test is strongly correlated with malignancy in mucinous cystic
lesions, particularly in recent multi-institutional molecular registry studies.
Future pancreatic cyst characterization in cases undiagnosed by cytopa-
thology will likely be largely dictated by cyst fluid analysis. Table 5.1 indi-
cates a number of DNA and biochemical markers, respectively, at various
stages of commercial development in the characterization of pancreatic
cysts. Importantly, CA19-9 and amylase have not been found to be accurate
indicators of malignancy or ductal connectivity.
MANAGEMENT
Nearly all decisions regarding the management of pancreatic cysts hinge on
either the presence of symptoms or the risk of existing or future progression
to malignancy. This highlights the critical role of clinical history, preopera-
tive imaging, and cytopathologic analysis in the evaluation and manage-
ment of these lesions. Although algorithms for preoperative evaluation and
operative decision making exist, the approach to each patient should be
individualized based on symptoms, anticipated malignant potential, extent
of disease, and patient fitness.
Operative Indications and Technical Tips
The indications for resection of cystic pancreatic lesions are typically
approached based on two primary considerations, symptoms and oncologic
risk. When dealing with lesions on the low end of the malignant risk spec-
trum (e.g., SCN), the decision to embark on pancreatic resection is driven
by the goal to alleviate present and future symptoms. For example, some
authorities have suggested that SCN should be considered for resection when
they reach a diameter of 4 cm or greater, as these lesions may have a greater
propensity for future growth and worsening symptoms. This preemptive
approach is not endorsed by all pancreatic surgeons, however; so patients
should be carefully considered for operative versus expectant management.
A much more clear set of indications are apparent in those patients
with evidence of invasive malignancy. These patients should be approached
with resection criteria similar to those utilized in pancreatic ductal ade-
nocarcinoma. Importantly, invasive pancreatic cystic lesions have dem-
onstrated a lower rate of vascular and lymphatic invasion than “garden
variety” pancreatic cancer, so are in general more amenable to resection.
The most controversial (and most studied) set of indications regarding
resection for cystic pancreatic lesions are for IPMN. Numerous retrospec-
tive series have demonstrated MPD involvement as an independent pre-
dictor of malignancy in IPMN (up to 60% at the time of resection). For this
002086374.INDD 58 7/14/2014 2:51:53 PM

reason, all fit patients with main duct involvement should be c onsidered
for pancreatic resection. Typically, the degree of oncologic risk in main
duct IPMN is defined based on the maximal segmental cystic dilation of
the MPD. An MPD ≥5 mm is worrisome, and if ≥10 mm, it is considered
a high-risk stigmata of malignancy according to the international consen-
sus guidelines (2012). The decision to resect cystic lesions not involving the
main duct is more complex. In surgical series, branch duct IPMN malig-
nant risk has been reported to occur in 5% to 20% of (mostly symptomatic)
patients. The incidence of malignancy in MCN is similar to that in IPMN.
The presence of high-risk stigmata (i.e., mural nodularity), concerning cyto-
pathology or multiple mutations on molecular profiling, should strengthen
the case for resection. The relationship between cyst size and malignancy
has been a debated topic in the literature, with most current series dem-
onstrating no correlation, though some demonstrate positive relationship.
Current international consensus guidelines no longer dictate a size cutoff
recommendation for operative resection.
When considering pancreatic resection in the management of cystic neo-
plasms of the pancreas, the primary goal is to perform segmental pancre-
atic resection to encompass all evident disease. In dealing with the unifocal
cystic lesions of the pancreas, this approach is usually straightforward. For
lesions of the pancreatic head and uncinate process, pancreaticoduode-
nectomy is the most common approach, whereas for lesions of the body/
tail, distal (left-sided) pancreatectomy is typically appropriate. In distal
pancreatectomy, splenic preservation is preferred unless there is a signifi-
cant chance invasive disease is present or the lesion is intimately associated
with the splenic hilum or vessels. Patients with larger (> 2.5 cm) cystic pNET
may also benefit from splenectomy for optimal lymph node retrieval. Some
authors have advocated enucleation in select cases and demonstrated
reduced operative morbidity and potential mortality. Enucleation should
only be considered if the preoperative likelihood of invasive malignancy is
expected to be low and the lesion is not in close proximity to or involving
the MPD. Tail lesions as they approach the spleen may be less optimal for
enucleation. This is in part due to minimal parenchyma spared with enucle-
ation and an elevated chance for MPD injury due to the small size of the
MPD and the difficulty of intraoperative ultrasound to accurately discrimi-
nate relationship of the cyst to the MPD. Multifocal cystic lesions, that is,
IPMN, may require substantially more complex preoperative planning. It is
not uncommon to have radiologic evidence of branch duct IPMN in mul-
tiple anatomic distributions, which would not be encompassed by a single
segmental pancreatic resection. The use of total pancreatectomy carries an
unbalanced and generally unacceptable risk of morbidity and mortality in
the setting of often premalignant and indolent disease. Thus, the approach
(except in unusual cases) is to perform segmental resection to include the
most oncologically concerning lesion(s) (i.e., mural nodularity, high-grade
atypia, main duct involvement), with subsequent surveillance of the resid-
ual gland/cysts.
Intraoperative Considerations
Several important technical points should be considered when conduct-
ing pancreatic resection for cystic lesions of the pancreas. Although cross-
sectional imaging may provide an excellent roadmap to guide the extent
of pancreatic resection, these lesions may be difficult to identify intraop-
eratively. Therefore, liberal use of intraoperative ultrasound to identify the
cyst or cysts should be considered, particularly when planning e nucleation,
0002086374.INDD 59 7/14/2014 2:51:53 PM

as this may aid in defining the anatomic relationship with the MPD.
Additionally, when undertaking resection for IPMN, particularly main duct
involved, frozen section should be obtained to ensure that no high-grade
dysplasia or invasive IPMN is present at the resection margin. Additional
margins should be taken when feasible in the setting of high-grade dyspla-
sia or frank invasion. If the main duct margin is positive for low- to mod-
erate-grade dysplasia but the gross lesion appears to be removed, further
resection is dictated based upon preoperative discussions of future symp-
tom management, malignant risk, patient preoperative pancreatic func-
tional reserve, and patient fitness.
Postoperative Management and Complications

The postoperative management of a patient undergoing segmental pan-
creatic resection for any cystic lesion of the pancreas is largely similar to
a patient undergoing pancreatectomy for other indications. Patients with
pancreatic cystic lesions typically have softer pancreatic parenchyma
(unless they have a history of significant pancreatitis). The MPD diameter
is also typically small, as there is rarely an obstructing mass. Given these
factors, the rate of pancreatic fistula has been reported to be elevated when
compared to fistula rates after resection for other indications. Additionally,
although enucleation may represent a less morbid approach to certain cys-
tic lesions, some authors have suggested an elevated risk of pancreatic fis-
tula development compared to segmental resection. Early recognition and
percutaneous drainage (if necessary) of pancreatic fistulas not controlled
by operatively placed drains (if present) is imperative to reduce major com-
plication such as systemic inflammatory response or sepsis, vascular pseu-
doaneurysm, breakdown of enteric or biliary reconstruction, and wound
infection/dehiscence.
OUTCOMES AND FOLLOW-UP

Cystic neoplasms of the pancreas represent a unique opportunity for the
clinician to intervene prior to the development of pancreatic malignancy.
When resected in the premalignant or in situ setting, the disease-specific
survival for IPMN or MCN may approach 95% to 100% at 5 years. Even in the
setting of invasive IPMN, the overall survival and stage-matched survival is
favorable when compared to patients undergoing resection for typical duc-
tal adenocarcinoma.
Patients undergoing resection for cystic lesions should undergo rou-
tine postoperative follow-up with clinical exam and cross-sectional imag-
ing as appropriate. For invasive cystic lesions, this will likely entail every
a 6-month exam and imaging surveillance unless dictated differently by
medical oncology treatment or trial protocol. For nonmalignant pNET and
SPN, this may entail interval exam and surveillance imaging for a period of
at least 5 years. For SCN or noninvasive MCN, as long as margins were clear,
surveillance imaging is not required. On the other hand, IPMN requires a
more involved postoperative surveillance strategy. As mentioned previ-
ously, because of its multicentric/multifocal nature, operative management
of IPMN often requires segmental resection of the most concerning lesion,
thereby leaving low-risk disease in the remnant gland. The risk profile of the
residual disease has been demonstrated to be related to the histologic grade
of the resected disease. Those patients with low- to moderate-grade IPMN
on initial resection have an exceedingly small risk of progression to malig-
nancy in the remnant gland, whereas patients with high-grade disease in
the resected specimen (even with negative margins) are at higher risk of
cancer progression in the remnant gland and should be considered for a
002086374.INDD 60 7/14/2014 2:51:53 PM

more intensive surveillance approach. Even patients without evidence of

residual disease in the remnant gland are at risk of de novo development
of IPMN or progression of disease that was radiologically occult at the time
of resection. For these reasons, surveillance with cross-sectional imaging
should be undertaken at routine intervals. Finally, numerous reports have
described an elevated risk of concomitant gastrointestinal malignancy in
patients with MCN or IPMN of the pancreas. These patients should be vigi-
lant with the American Cancer Society recommendations about screening
for prostate, colon, breast, and lung cancer. The clinician should maintain a
high index of suspicion for nonpancreatic neoplasia in these patients.
CONCLUSIONS
Cystic neoplasms of the pancreas are an increasingly recognized entity
with important clinical implications. The accurate diagnosis and charac-
terization of these lesions with clinical history, cross-sectional imaging (CT,
MRCP, EUS), as well as cytopathologic sampling are critical in oncologic
risk stratification and subsequent clinical decision making. The subset of
cysts with little to no malignant potential should only be considered for
resection in the setting of substantial size, mass effect, or refractory symp-
toms. Cysts with malignant features or high malignant potential should
be considered for operative resection in fit patients. Primary surveillance
strategies may be acceptable for low-risk lesions, as pancreatic resection
carries substantial morbidity. Surveillance of the pancreatic remnant after
resection is dictated based on cyst diagnosis and dysplastic grade. In the
setting of IPMN, surveillance should be conducted indefinitely as these
lesions are commonly multicentric and may recur or develop de novo in
the remnant gland.
Suggested Readings
Cauley CE, Waters JA, Dumas RP. Outcomes of primary surveillance for intraductal
papillary mucinous neoplasm. J Gastrointest Surg 2012;16(2):258–267; discussion
266.
D’Angelica M, Brennan MF, Suriawinata AA, et al. Intraductal papillary mucinous neo-
plasms of the pancreas: an analysis of clinicopathologic features and outcome.
Ann Surg 2004;239:400.
Katabi N, Klimstra DS. Intraductal papillary mucinous neoplasms of the pan-
creas: clinical and pathologic features and diagnostic approach. J Clin Pathol
2008;61;1303–1313.
Schmidt CM, White PB, Waters JA, et al. Intraductal papillary mucinous neoplasms:
predictors of malignant and invasive pathology. Ann Surg 2007;246(4):644–654.
Sohn TA, Yeo CJ, Cameron JL, et al. Intraductal papillary mucinous neoplasms of the
pancreas: an updated experience. Ann Surg 2004;239:788–797.
Tanaka M, Fernández-del Castillo C, Adsay V, et al. International Consensus Guidelines
2012 for the Management of IPMN and MCN of the Pancreas. Pancreatology
2012;12(3):183–197.
Wargo JA, Fernandez-del-Castillo C, Warshaw AL. Management of pancreatic serous
cystadenomas. Adv Surg 2009;43:23–34.
Waters JA, Schmidt CM. Intraductal papillary mucinous neoplasm—when to resect?
Adv Surg 2008;42:87–108.
0002086374.INDD 61 7/14/2014 2:51:53 PM

6 Pancreatic Neuroendocrine
Neoplasms
Rebecca A. Busch and Clifford S. Cho
INTRODUCTION
Pancreatic neuroendocrine neoplasms are relatively rare. The incidence of clin-
ically detected cases is approximately 4 cases per million people per year in the
United States. Like the more common exocrine pancreatic adenocarcinoma,
they are capable of regional and distant metastatic dissemination. But, they
can also be relatively indolent. Indeed, autopsy studies suggest that the actual
incidence of pancreatic neuroendocrine neoplasms may be as high as 1.5%.
This combination of relative rarity and indolence has challenged the ability of
investigators to definitively characterize the natural history and optimal treat-
ment of this disease. However, recent studies to be outlined in this chapter have
yielded important new insights for the hepato-pancreatico-biliary surgeon.
ETIOLOGY
Initially believed to arise from the endocrine islets of Langerhans, recent
studies suggest that these neoplasms originate from pluripotent cells
within pancreatic ductules. As a result, previous descriptors like “islet cell
tumors” and “islet cell carcinomas” have been dropped in favor of “pancre-
atic neuroendocrine neoplasms.” Pancreatic neuroendocrine neoplasms
have been categorized into “functioning” variants (those that overproduce
pancreatic endocrine hormones capable of inducing specific clinical signs
and symptoms) and “nonfunctioning” variants (those that exhibit no clear
pattern of symptomatic hormonal overproduction).
Pancreatic neuroendocrine neoplasms can occur sporadically or as a
component of inherited genetic syndromes. Genetic and epigenetic altera-
tions of the p16/MTS1 tumor suppressor gene have been implicated in the
development of sporadic pancreatic neuroendocrine neoplasms. Whereas
loss of heterozygosity at chromosome 11q appears to be associated with
functional lesions, loss of heterozygosity at chromosome 6q has been asso-
ciated with nonfunctional neoplasms. The inherited genetic syndrome
most commonly associated with pancreatic neuroendocrine neoplasms
is multiple neuroendocrine neoplasia type 1 (MEN 1), an autosomal domi-
nant trait associated with mutations of the presumed tumor suppressor
gene menin. In addition, von Hippel-Lindau disease and neurofibromatosis
type 1 are also associated with pancreatic neuroendocrine neoplasms.
DIAGNOSIS
Clinical Presentation
Historically, it was believed that functioning variants comprised the major-
ity of pancreatic neuroendocrine neoplasms; however, as more diagnoses
are being made by imaging and not by symptoms, most contemporary
series suggest that nonfunctioning neoplasms comprise the large majority.
Nevertheless, the often bizarre and characteristic behavior of functional
neoplasms warrants discussion of their various subtypes (Table 6.1).
62
002086375.INDD 62 7/14/2014 3:04:59 PM

0002086375.INDD 63
TABLE
Summary of Functioning Pancreatic Neuroendocrine Neoplasms
6.1
Neoplasm Common Clinical Characteristics Secretory Product Symptoms Diagnostic Criteria
Insulinoma Small (< 2 cm) Insulin Whipple triad: During 72-hour fast:
90% solitary (promotes hypoglycemia) 1. Hypoglycemic symptoms Glucose < 50 mg/dL
Evenly distributed throughout during monitored fast Insulin > 5 μ/mL
pancreas 2. Blood glucose < 50 mg/dL Insulin: glucose ratio > 0.4
90% benign with symptoms Elevated C-peptide, proinsulin
Seen in 5%–8% of MEN1 3. Symptom relief after admin-
istration of glucose
Gastrinoma Sporadic form: solitary Gastrin Zollinger-Ellison syndrome: Gastrin > 10 × upper limit normal
MEN1: multifocal (promotes gastric acid 90% gastroduodenal Gastrin increase > 200 pg/mL
90% within gastrinoma triangle production) ulcerations after secretin stimulation
50% malignant 75% epigastric pain
30% present with metastases 75% diarrhea
15%–35% associated with MEN1
Glucagonoma Large Glucagon “4 Ds”: Fasting glucagon > 1,000 pg/mL
Solitary (promotes hyperglycemia, Diabetes Skin biopsy of necrolytic
Body/tail of the pancreas amino acid metabolism) Dermatitis (necrolytic migratory erythema
75% malignant migratory erythema)
Deep venous thrombosis
Depression
(Continued)
Chapter 6 / Pancreatic Neuroendocrine Neoplasms 63
7/14/2014 3:04:59 PM
002086375.INDD 64
TABLE
Summary of Functioning Pancreatic Neuroendocrine Neoplasms (Continued)
6.1
Neoplasm Common Clinical Characteristics Secretory Product Symptoms Diagnostic Criteria
VIPoma Large Vasoactive intestinal Verner-Morrison/WDHA Fasting VIP > 500 pg/mL

Solitary polypeptide (VIP) syndrome:
Body/tail of the pancreas (promotes intestinal Watery diarrhea
Malignant secretion and motility Hypokalemia
50% present with metastases and inhibits intestinal Achlorhydria/hypochlorhydria
absorption)
Somatostatinoma Large Somatostatin Diabetes Fasting somatostatin > 100 pg/mL
Solitary (inhibits pancreatic Malabsorption
Head of the pancreas endocrine and exocrine Cholelithiasis
Malignant secretion and gallbladder
Commonly present with contractility)
metastases
Associated with von
Recklinghausen disease
(NF type 1)
7/14/2014 3:04:59 PM
Insulinomas are the most common type of functional pancreatic

neuroendocrine neoplasm. They tend to be small (< 2 cm), solitary, benign
lesions that can arise anywhere within the pancreas and can be associated
with MEN1 (although less commonly than gastrinomas). The diagnosis
of insulinoma can be made by the classic Whipple triad of clinical find-
ings: (1) symptoms of hypoglycemia during monitored fasting or exercise;
(2) blood glucose levels of less than 50 mg/dL during these symptomatic epi-
sodes; and (3) relief of these symptoms following administration of glucose.
To detect the Whipple triad, patients with suspected insulinoma are closely
monitored during a 72-hour inpatient fast. Following measurement of base-
line serum glucose and insulin levels, blood glucose levels are measured
every 2 hours, and insulin levels are measured at the onset of hypoglyce-
mic symptoms or when blood glucose levels become lower than 50 mg/dL.
Nearly all patients with insulinoma will have inappropriately elevated
plasma insulin levels (> 5 μU/mL) in the setting of hypoglycemia. Elevated
levels of C peptide and proinsulin are also typically present and help to dif-
ferentiate insulinoma from factitious hypoglycemia; the onset of symptoms
during fasting helps to differentiate insulinoma from postprandial reactive
hypoglycemia.
Gastrinomas are the second most common type of functional pancre-
atic neuroendocrine neoplasm. Unlike insulinomas, 90% of gastrinomas
are found within the gastrinoma triangle: an anatomical region defined by
the cystic and common bile ducts superiorly, the second and third portions
of the duodenum inferiorly, and the pancreatic neck and body medially
(Fig. 6.1). Two-thirds of gastrinomas are located within the pancreas, and a
third may be found within the duodenum. In sporadic, noninherited forms
of gastrinoma, tumors tend to be solitary. However, gastrinomas arising in
the context of MEN1 are multifocal and highly prone to recurrence. Because
of the ability of gastrin to induce gastric acid secretion, gastrinomas typi-
cally present with the classic and often refractory signs and symptoms of
Zollinger-Ellison syndrome, consisting of gastroduodenal ulcerations (90%),
epigastric pain (75%), diarrhea associated with the large-volume gastric
acid secretion (75%), and malabsorption leading to weight loss. The inci-
dence of gastrinoma among patients presenting with peptic ulcer disease
has been estimated to be approximately 1%, and this diagnosis should be
entertained in patients presenting with advanced or refractory peptic ulcer
disease. Unlike insulinomas, gastrinomas are often malignant, with about
50% of patients presenting with evidence of hepatic metastases.
Patients with suspected gastrinoma should undergo serum gastrin
measurement. The upper limit of normal gastrin level is approximately
100 pg/mL; serum gastrin levels ten times greater than the upper limit
of normal in the presence of gastric pH less than 5.0 are diagnostic of
gastrinoma. However, patients with gastrinoma often present with more
subtle degrees of hypergastrinemia. The differential diagnosis for mod-
erate hypergastrinemia includes achlorhydria (as can occur with proton
pump inhibitor therapy), retained antrum following partial gastrectomy,
gastric outlet obstruction, and renal insufficiency. The secretin stimula-
tion test is a means of increasing the sensitivity of detecting gastrinoma.
Following intravenous administration of 2 μg/kg secretin, a paradoxical
rise in serum gastrin by greater than 200 pg/mL is considered positive
for gastrinoma. Similarly, a paradoxical rise in serum gastrin of greater
than 50% following infusion of calcium gluconate is also indicative of
gastrinoma.
Unlike insulinomas and gastrinomas, glucagonomas are characteristi-
cally larger lesions that tend to be localized along the body and tail of the
pancreas. Like gastrinomas, they exhibit a proclivity for aggressive behavior,
0002086375.INDD 65 7/14/2014 3:04:59 PM

FIGURE 6.1 Most gastrinomas are found within the gastrinoma triangle. (From: Mulholland
MW, Lillemoe KD, Doherty GM, et al. Greenfield’s surgery: scientific principles & practice,
5th ed. Philadelphia, PA: Lippincott Williams & Wilkins, 2011.)
with approximately 50% of cases harboring evidence of distant metastatic

disease. The ability of glucagon to promote gluconeogenesis and amino acid
oxidation explains the characteristic clinical findings of diabetes (resulting
from hyperglycemia) and dermatitis (resulting from amino acid deficiency).
Skin manifestations of dermatitis include necrolytic migratory erythema,
characterized by painful blistering plaques along the face, abdomen, lower
extremities, and mucous membranes. Interestingly, up to a third of patients
with glucagonoma develop deep venous thrombosis, warranting consider-
ation of anticoagulation therapy.
Serum glucagon levels greater than 1,000 pg/mL can be diagnostic
for glucagonoma. However, as with gastrinoma, more subtle elevations
can be seen in patients with glucagonoma. Unlike gastrinoma, provocative
testing is not available to help distinguish glucagonoma from potentially
confounding conditions that can induce hyperglucagonemia such as pan-
creatitis, sepsis, Cushing syndrome, fasting, renal insufficiency, and hepatic
failure.
Like glucagonomas, VIPomas often present as large, solitary lesions
localized to the pancreatic body and tail and often present with evidence
of metastatic disease. Rare extrapancreatic VIPomas have been observed
in the colon, bronchus, liver, adrenal glands, and para-aortic ganglia.
Vasoactive intestinal polypeptide (VIP) promotes intestinal secretion and
002086375.INDD 66 7/14/2014 3:05:01 PM

motility while inhibiting intestinal absorption of water and electrolytes.

As a result, the classic Verner-Morrison syndrome of VIPomas is also referred
to by the acronymic WDHA syndrome (consisting of massive watery diar-
rhea, hypokalemia, and achlorhydria). In patients presenting with profuse
secretory diarrhea, the diagnosis of VIPoma can be made by measuring
elevated fasting serum VIP levels (typically > 500 pg/mL).
In contrast to glucagonomas and VIPomas, the majority of somatostati-
nomas are found within the pancreatic head. Like gastrinomas and
VIPomas, somatostatinomas can present outside of the pancreas (most
commonly in the duodenum and ampulla). Duodenal somatostatinomas
are a common manifestation of von Recklinghausen disease (neurofibroma-
tosis type 1). Pancreatic somatostatinomas tend to present as larger tumors
with metastatic disease. Somatostatin inhibits secretion of pancreatic insu-
lin and exocrine enzymes and inhibits gallbladder contractility; as a result,
somatostatinomas often present with symptoms of diabetes, malabsorp-
tion, and cholelithiasis. The diagnosis of somatostatinoma among patients
presenting with these symptoms is supported by elevated serum levels of
somatostatin greater than 100 pg/mL.
Nonfunctioning pancreatic neuroendocrine neoplasms tend to pres-
ent evenly throughout all regions of the pancreas. Taken as a whole, they
were previously characterized as generally rarer, larger, and more aggres-
sive than functioning pancreatic neuroendocrine neoplasms. However,
these characterizations may have been historical artifacts of diagnosis.
As more and more pancreatic neuroendocrine neoplasms are being diag-
nosed incidentally with imaging evaluations rather than by symptoms
referable to hormonal production or tumor mass, the prevalence of small,
asymptomatic, nonfunctioning pancreatic neuroendocrine neoplasms
appears to be increasing. Thus, the extent to which functioning and non-
functioning pancreatic neuroendocrine neoplasms differ in terms of inher-
ent tumor biology and behavior is unclear. Nonfunctioning lesions can
occasionally secrete measurable levels of proteins that do not induce sig-
nificant symptomatology, but often do not express any known hormones.
As a result, elevated serum levels of proteins like pancreatic polypeptide,
neuron-specific enolase, and neurotensin can have high specificity but
very poor sensitivity in the diagnosis of nonfunctioning pancreatic neuro-
endocrine neoplasms.
Imaging and Localizing Studies

Diagnostic imaging for pancreatic neuroendocrine neoplasms takes advan-
tage of some of their unique biologic characteristics. The majority of pancre-
atic neuroendocrine neoplasms are markedly well vascularized. As a result,
contrast-enhanced imaging by computed tomography (CT) or magnetic
resonance imaging (MRI) should employ both early arterial and delayed
portal-phase imaging sequences, as they tend to appear very bright during
early arterial-phase administration of contrast (Fig. 6.2) or on portal-phase
imaging (Fig. 6.3). Radiographic diagnosis can be further enhanced with
MRI, as pancreatic neuroendocrine neoplasms tend to exhibit low signal
intensity on T1-weighted imaging and high signal intensity on T2-weighted
imaging (Fig. 6.4). These imaging characteristics can also be used in the
detection of extrapancreatic metastases.
Another relevant biologic characteristic of pancreatic neuroendocrine
neoplasms is that they tend to grow as rounded, well-circumscribed lesions.
The ability of endoscopic ultrasonography (EUS) to clearly recognize these
morphologic characteristics (Fig. 6.5) may explain its heightened sensitiv-
ity in the detection of smaller lesions as compared with cross-sectional CT
and MRI. Intraoperative ultrasonography with direct probe application on
0002086375.INDD 67 7/14/2014 3:05:01 PM

FIGURE 6.2 CT imaging of neuroendocrine neoplasm arising from the uncinate process
demonstrates brisk uptake of contrast on early arterial-phase imaging (white arrow).
the surface of the pancreas or duodenum can be especially useful in the

detection of occult and multifocal lesions (as can be seen with insulinomas
or with duodenal or pancreatic gastrinomas arising in a setting of MEN1).
Somatostatin receptor scintigraphy (octreotide scanning) takes
advantage of the fact that many pancreatic neuroendocrine neoplasms
FIGURE 6.3 CT imaging of neuroendocrine neoplasm arising from the pancreatic head
demonstrates contrast enhancement on portal-phase imaging with areas of central necrosis
(white arrow).
002086375.INDD 68 7/14/2014 3:05:02 PM

FIGURE 6.4 A. MRI imaging of neuroendocrine neoplasm arising from the pancreatic body
demonstrates a rounded lesion of low signal intensity on T1-weighted sequences (white
arrow). B. MRI imaging of the same neuroendocrine neoplasm arising from the pancreatic
body demonstrates high signal intensity on T2-weighted sequences (white arrow).
FIGURE 6.5 EUS imaging of the neuroendocrine neoplasm arising from the pancreatic tail
demonstrates a rounded and well-circumscribed lesion (black arrow).
0002086375.INDD 69 7/14/2014 3:05:03 PM

overexpress somatostatin receptor proteins. Octreotide is a s omatostatin

analogue (and somatostatin receptor ligand) that can be radiolabeled
with 111In and administered intravenously. Scintigraphic localization of
radiolabeled octreotide uptake, especially when used in conjunction with
single-photon emission computed tomography, can identify both primary
pancreatic neuroendocrine neoplasms and foci of metastatic disease with
high sensitivity. Recent analyses, however, suggest that they rarely yield
additional information beyond that which can be obtained with traditional
cross-sectional imaging studies alone. Of note, insulinomas typically do
not overexpress somatostatin receptor and are thus poorly visualized with
somatostatin receptor scintigraphy.
Venous sampling is a means of tumor localization that takes advantage
of pancreatic neuroendocrine neoplasm’s ability to secrete hormones into
the portal venous circulation. Now largely of historic interest, venous sam-
pling involves the placement of percutaneous transhepatic catheters into
the portal venous circulation. These catheters can be positioned within the
portal vein, splenic vein, or superior mesenteric vein and are used to collect
venous blood samples for insulin or gastrin measurement. Identifying the
venous basin in which hormone levels are elevated may facilitate the local-
ization of radiographically occult insulinomas and gastrinomas. The utility
of this technique is enhanced with the addition of selective arterial stimula-
tion. In this approach, a percutaneous transarterial catheter is positioned
within the common hepatic artery, splenic artery, or gastroduodenal artery
and used for selective infusion of either secretin or calcium. The ability of
these agents to stimulate hormonal production can then be used to localize
tumors to various locations (e.g., pancreatic tail or body vs. pancreatic head
or neck, hepatic metastases) based on their arterial supply.
MANAGEMENT
Until recently, safe and effective chemotherapeutic agents were not avail-
able for the treatment of pancreatic neuroendocrine neoplasms. As a
result, surgical resection has been the mainstay of therapy for this disease.
Complete surgical resection remains the only potentially curative inter-
vention for pancreatic neuroendocrine neoplasms. However, because the
course of this disease can be gradual and indolent, operative intervention
has also been employed in the attempt to palliate symptoms and possibly
prolong survival for patients with metastatic disease.
Localized Disease
Surgical resection is the standard treatment intervention for patients
with localized and resectable pancreatic neuroendocrine neoplasm. The
specific type of resection is tailored to the location and distribution of
disease and to the patient’s ability to tolerate a major pancreatic resec-
tion. The standard operative approach for resection of tumors within
the pancreatic head or uncinate process is pancreaticoduodenectomy,
whereas left or distal pancreatectomy is reserved for tumors within the
pancreatic body or tail. Total pancreatectomy is typically reserved for
large or multifocal tumors involving the pancreatic head, neck, and
body. Central pancreatectomy has been advocated as a means of pre-
serving functional pancreatic parenchyma for patients with tumors
localized to the pancreatic neck or body; however, any potential advan-
tage in endocrine function must be counterbalanced by the risk of fistula
at the site of the pancreaticojejunostomy or pancreaticogastrostomy
that must be constructed to drain the pancreatic tail remnant. For small
tumors in locations not involving the main pancreatic duct, enucleation
002086375.INDD 70 7/14/2014 3:05:03 PM

has been advocated as an alternative and potentially less invasive

means of tumor extirpation. In this technique, dissection p roceeds
along the peritumoral plane, and only the tumor is removed. Because
of the limited extent of resection (as compared with partial or total
pancreatectomy), enucleation carries a higher risk of positive resection
margins and does not permit pathologic assessment of peripancreatic
lymph nodes. Therefore, although enucleation has been shown to be
associated with lower operative morbidity, this potential advantage
must be counterbalanced by the risk of oncologically incomplete tumor
clearance. Surgical margins and nodal metastases have been shown to
be prognostically significant variables for patients with resected pan-
creatic neuroendocrine neoplasms; therefore, enucleation is typically
reserved for patients with tumors that are very likely to be benign (e.g.,
insulinomas) or for patients who may not tolerate radical resection. In
this context, retrospective studies have not consistently identified mea-
surable survival differences between patients treated with enucleation
versus radical resection.
Gastrinoma Arising in a Setting of MEN1

One circumstance that deserves special attention is that of gastrinoma
arising in the setting of MEN1. These patients harbor a genetic predis-
position toward the development of multifocal gastrinoma, as mani-
fested by their very high rates of local and distant disease recurrence
following surgical intervention. In fact, the goal of surgical intervention
in these patients is often directed more at control of refractory peptic
ulcer disease–related symptoms than at cure. Indeed, whether and how
to undertake operative therapy for patients with gastrinoma and MEN1
remains an area of controversy and debate. Preoperatively, every effort
is expended to define the extent of disease within the pancreas and peri-
pancreatic tissues using cross-sectional CT and/or MRI modalities as
well as EUS, as utilization of multiple techniques maximizes the sensitiv-
ity of multifocal tumor detection. However, because of finite limits in the
sensitivity of even these preoperative maneuvers, additional maneuvers
are undertaken at the time of operative exploration in order to maximally
define the extent of disease. In addition to performing intraoperative US
of the pancreas and peripancreatic tissues, many surgeons advocate
performing a duodenotomy to more fully evaluate for small tumors pos-
sibly residing within the duodenal wall. Digital palpation and transillu-
mination of the entire duodenal wall have been suggested as effective
adjuncts to enhance the detection of small tumor foci. Operative resec-
tion should then be tailored toward the distribution of disease that is
present, and often necessitates a distal pancreatectomy with resection of
any duodenal wall tumors that may be present. On occasion, a total pan-
createctomy may be necessary. Because of the high likelihood of nodal
metastases, a regional peripancreatic lymphadenectomy should also be
undertaken.
Metastatic Disease
Like other gastroenteropancreatic malignancies, neuroendocrine neo-
plasms have a tendency to metastasize to the liver. Treatment directed
at hepatic neuroendocrine metastases has been advocated as a means of
mitigating tumor burden–mediated symptoms and, possibly, as a way to
prolong survival.
In well-selected cases, surgical resection of hepatic neuroendo-
crine metastases has been associated with favorable control of symp-
toms and with durable survival outcomes. However, these analyses have
0002086375.INDD 71 7/14/2014 3:05:03 PM

been limited to noncontrolled, retrospective studies, some of which

have suboptimally compared patients undergoing surgical resection to
patients who were not candidates for operative intervention. Analyses
that have included large patient numbers and durable follow-up sug-
gest that the likelihood of eventual disease recurrence following surgical
metastasectomy is very high. Given the relative rarity and the relatively
indolent disease course of even metastatic neuroendocrine neoplasms,
the extent to which hepatic metastasectomy may prolong survival
remains unclear.
Hepatic neuroendocrine metastases are commonly multifocal and
are often not amenable to complete operative resection. In many cases,
operative resection is combined with intraoperative microwave or radio-
frequency ablation in an effort to preserve hepatic parenchyma. Some
investigators have suggested that even subtotal cytoreduction can offer
palliation of symptoms and prolongation of survival for patients with
extensive hepatic neuroendocrine metastases. For many patients with
extensive multifocal disease, liver-directed therapy relies on transarterial
chemoembolization or radioembolization. These techniques take advantage
of the fact that hepatic metastases derive the majority of their blood sup-
ply from the hepatic arterial circulation (as opposed to the portal venous
circulation). Selective administration of chemotherapy (most commonly
doxorubicin and cisplatin) or radioactive microspheres is followed by arte-
rial embolization. Transarterial embolization therapy has been shown
to be capable of controlling symptoms and inducing tumoral regression,
but prospective randomized demonstration of survival benefit is not yet
available.
Systemic Therapy
Until recently, systemic therapy for patients with pancreatic neuroendo-
crine neoplasms was restricted to somatostatin analogs and cytotoxic che-
motherapy. Unfortunately, neither approach offered significant efficacy.
Somatostatin analogs like octreotide or the long-acting variant lanreotide
are able to inhibit hormonal secretion. In addition, they have been shown
to inhibit proliferation of an insulinoma cell line in vitro. However, they
do not appear to have any direct cytotoxic effect. Consequently, soma-
tostatin analogue therapy has been shown to be effective in ameliorat-
ing hormonal symptoms in 60% to 90% of patients; however, patients
typically develop resistance to this therapy after about 1 year, and actual
induction of tumoral regression is rare (5% to 15%). Cytotoxic chemo-
therapy using 5-fluorouracil, streptozocin, and doxorubicin has been
shown to have a response rate of 39%. A combination of cisplatin and
etoposide has been shown to be associated with a response rate of 42%
among patients with poorly differentiated tumors. However, both these
regimens are associated with significant risks of adverse effects that
often limit their utility.
Very recently, targeted therapies used in the treatment of other
malignancies have been shown to have efficacy against pancreatic
neuroendocrine neoplasms. Moreover, these agents are associated with
comparatively minimal side effects. Two of these agents, sunitinib and
everolimus, have recently received U.S. FDA approval for use in the treat-
ment of pancreatic neuroendocrine neoplasms. Sunitinib is a tyrosine
kinase inhibitor that targets a number of angiogenic and mitogenic pro-
teins including vascular endothelial growth factor receptor and platelet-
derived growth factor receptors. In a recent phase 3 randomized clinical
trial, treatment with sunitinib promoted significant prolongation of median
002086375.INDD 72 7/14/2014 3:05:03 PM

rogression-free survival (11.4 months vs. 5.5 months) and significantly higher
p
objective response rates (9% vs. 0%) as compared with placebo for patients
with well-differentiated pancreatic neuroendocrine neoplasms. Everolimus
is an inhibitor of the protein mammalian target of rapamycin. By inhibiting
the cellular proliferation and angiogenic events mediated through mTOR
signaling, everolimus can inhibit the growth and survival of pancreatic neu-
roendocrine neoplasm cell lines in vitro. In a recent multicenter randomized
clinical trial, everolimus induced significant prolongation of progression-
free survival (11 months vs. 5.4 months) and significantly higher objective
response rates (5% vs. 2%) compared with placebo for patients with well-
differentiated pancreatic neuroendocrine neoplasms. Importantly, both
sunitinib and everolimus were found to have acceptable side effect profiles.
Current investigations are examining newer targeted therapy agents as well
as combination therapies.
OUTCOMES
Retrospective reports of patients undergoing complete resection of pan-
creatic neuroendocrine neoplasms are relatively heterogenous in terms of
their treatment approaches and patient characteristics. However, review
of representative studies with reasonable patient numbers and follow-up
durations suggests that survival outcomes after resection may be gener-
ally favorable, with 5-year survival estimates ranging between 60% and
90% (Table 6.2). Prediction of individual patient outcomes relies on the
identification of prognostically informative clinical variables, and optimal
classification of pancreatic neuroendocrine neoplasms, which remains an
area of persistent controversy. The seventh edition of the American Joint
Commission on Cancer (AJCC) Cancer Staging Manual outlines a staging
system for pancreatic neuroendocrine neoplasms that follows a conven-
tional tumor–node–metastasis (TNM) schema (Table 6.3). However, it has
been evident for some time that tumor grade, as quantified by mitotic activ-
ity, carries significant prognostic weight for neuroendocrine neoplasms.
The World Health Organization (WHO) refined their classification system
in 2010 to accommodate this variable (Table 6.4). Although histologic
grade does stratify prognostic outcomes, a number of series have shown
that variables such as tumor size, nodal metastases, and resection margin
are prognostically informative variables for patients with resected pancre-
atic neuroendocrine neoplasms. As a result, optimal staging of this disease
is likely to undergo further evaluation and modification in the future.
TABLE
Summary of Recent Retrospective Analyses of
6.2 Patients Undergoing Surgical Resection of Pancreatic

Neuroendocrine Neoplasms
Median
Number of Follow-up 5-Year
Authors Year Patients (Months) Survival
Kazankian, et al. 2006 70 50 89%
Schurr, et al. 2007 45 55 64%
Ferrone, et al. 2007 183 44 87%
Bilimoria, et al. 2008 3,855 51 59%
Hill, et al. 2009 1,815 26 ~ 70%
Ballian, et al. 2009 43 68 91%
0002086375.INDD 73 7/14/2014 3:05:03 PM

002086375.INDD 74
TABLE
AJCC Classification System of Pancreatic Neuroendocrine Neoplasms
6.3
Primary tumor (T) Stage T N M
Tx Primary tumor cannot be assessed 0 Tis N0 M0
T0 No evidence of primary tumor IA Tl N0 M0

Tis Carcinoma in situ IB T2 N0 M0
T1 Tumor limited to pancreas, ≤ 2 cm in greatest dimension IIA T3 N0 M0
T2 Tumor limited to pancreas, > 2 cm in greatest dimension IIB Tl Nl M0
T3 Tumor extends beyond pancreas but without involvement of the celiac axis or T2 Nl M0
the superior mesenteric artery
T4 Tumor involves the celiac axis or the superior mesenteric artery (unresectable T3 Nl M0
primary tumor)
III T4 any N M0
Regional lymph nodes (N) IV any T any N Ml
Nx Regional lymph nodes cannot be assessed

N0 No regional lymph node metastases
N1 Regional lymph node metastases
Distant metastasis (M)
M0 No distant metastasis
M1 Distant metastasis
7/14/2014 3:05:04 PM
TABLE
WHO Classification of Pancreatic Neuroendocrine
6.4 Neoplasms
Ki-67 Staining
Differentiation Grade Mitotic Figures Positivity
Well differentiated Low grade (G1) <2 per 10 HPFa <3%
Intermediate 2 − 20 per 10 3%− 20%
grade (G2) HPFa
Poorly differentiated High grade (G3) >20 per 10 HPFa >20%
a
high-power fields
Suggested Readings
Ballian N, Loeffler AG, Rajamanickam V, et al. A simplified prognostic scoring
system for resected pancreatic neuroendocrine neoplasms. HPB (Oxford)
2009;11:422–428.
Bilimoria KY, Bentrem DJ, Merkow RP, et al. Application of the pancreatic adenocar-
cinoma staging system to pancreatic neuroendocrine tumors. J Am Coll Surg
2007;205:558–563.
Ferrone CR, Tang LH, Tomlinson J, et al. Determining prognosis in patients with pan-
creatic endocrine neoplasms: can the WHO classification system be simplified?
J Clin Oncol 2007;25:5609–5615.
Hill JS, McPhee JT, McDade TP, et al. Pancreatic neuroendocrine tumors: the impact of
surgical resection on survival. Cancer 2009;115:741–751.
Kazanjian KK, Reber HA, Hines OJ. Resection of pancreatic neuroendocrine tumor:
results of 70 cases. Arch Surg 2006;141:765–770.
Martin RC, Kooby DA, Weber SM, et al. Analysis of 6,747 pancreatic neuroendocrine
tumors for a proposed staging system. J Gastrointest Surg 2011;15:175–183.
Norton JA, Fraker DL, Alexander HR, et al. Surgery increases survival in patients with
gastrinoma. Ann Surg 2006;244:410–419.
Raymond E, Dahan L, Raoul JL, et al. Sunitinib malate for the treatment of pancreatic
neuroendocrine tumors. N Engl J Med 2011;364:501–513.
Schurr PG, Strate T, Rese K, et al. Aggressive surgery improves long-term survival in
neuroendocrine pancreatic tumors. Ann Surg 2007;245:273–281.
Yao JC, Shah MH, Ito I, et al. Everolimus for advanced pancreatic neuroendocrine
tumors. N Engl J Med 2011;364:514–523.
0002086375.INDD 75 7/14/2014 3:05:04 PM

7
Technical Aspects of
Pancreatic Surgery
Charles H. C. Pilgrim, Douglas B.
Evans, and Kathleen K. Christians
PREOPERATIVE CONSIDERATIONS
Pancreatic resection, in particular, pancreaticoduodenectomy, is a tech-
nical and perioperative tour de force in abdominal surgery and should be
performed only in carefully selected patients. As with any major abdominal
procedure, preoperative evaluation for pancreatic surgery should include
a detailed history and physical examination (including functional status),
laboratory studies, and optimization of medical comorbidities. Tumor
markers (serum carbohydrate antigen [Ca19-9], carcinoembryonic antigen
[CEA]) have prognostic value, and CA19-9 is being used to identify those
patients at very high risk for harboring subclinical, radiographically occult,
distant metastases. High-resolution abdominal computed tomography
(CT) imaging is the mainstay of clinical staging. Contrast-enhanced multi-
detector pancreas protocol CT provides important information regarding
tumor–vessel relationships. In particular, direct assessment of the relation-
ship to the major vascular structures (superior mesenteric artery [SMA],
superior mesenteric vein [SMV], SMV–portal vein confluence [SMV–PV],
celiac axis, and common hepatic artery [CHA]) is readily determined in a
reproducible manner. In addition, arterial and venous anomalies that affect
the technical aspects of the operation should be noted and potential sites
of metastatic disease identified. Resectability status must be established on
the basis of high-quality imaging prior to treatment as it forms the basis for
which stage-specific therapy is delivered both as part of a clinical trial and
as off-protocol therapy. Resectability criteria utilized by the Medical College
of Wisconsin pancreatic cancer team are found in Table 7.1.
PANCREATICODUODENECTOMY
Pancreaticoduodenectomy involves resection of the pancreatic head, duo-
denum, gallbladder, and bile duct with or without removal of the gastric
antrum. Our recommended technique includes a brief staging laparoscopy
to exclude liver and peritoneal metastases followed by a midline upper
abdominal incision.
The surgical resection is divided into the following steps:
1. Isolation of the infrapancreatic SMV and separation of the colon and
its mesentery from the duodenum and pancreatic head: The lesser sac
is entered by mobilizing the greater omentum up off of the transverse
colon. The loose attachments of the posterior gastric wall to the ante-
rior surface of the pancreas are divided. The hepatic flexure of the colon
is freed from its retroperitoneal attachments, exposing the pancreatic
head and duodenum. The visceral peritoneum along the inferior border
of the pancreas is incised starting from the left of the middle colic ves-
sels toward the patient’s right and inferiorly to expose the junction of
76
002086376.INDD 76 7/14/2014 3:15:32 PM

0002086376.INDD 77
TABLE
Medical College of Wisconsin CT-based Clinical Staging of Pancreatic Cancer
7.1 Tumor–Artery Relationship Tumor–Vein Relationship Extrapancreatic Disease
Resectable No radiographic evidence of Tumor-induced narrowing <50% of SMV, PV, Nil
arterial abutment (celiac, SMA, or SMV–PV
or hepatic artery)
Borderline Tumor abutment (<180 degrees) of Tumor-induced narrowing of >50% of SMV, PV, CT scan findings suspicious,
the SMA or celiac artery or SMV–PV confluence but not diagnostic of
Tumor abutment or short-segment Short-segment occlusion of SMV, PV, SMV–PV metastatic disease (e.g., small
encasement (>180 degrees) of with suitable PV (above), and SMV (below) indeterminate liver lesions that
the hepatic artery to allow for safe vascular reconstruction are too small to characterize)
Locally advanced Tumor encasement (>180 degrees) Occlusion of SMV, PV, or SMV–PV without No evidence of peritoneal, hepatic,
of the SMA or celiac artery suitable vessels above and below the tumor or extra-abdominal metastases
to allow for reconstruction (no distal or
proximal target for vascular reconstruction)
Metastatic N/A N/A Evidence of peritoneal or distant
metastases
SMA, superior mesenteric artery; SMV, superior mesenteric vein; PV, portal vein; SMV–PV, superior mesenteric–portal vein confluence
Chapter 7 / Technical Aspects of Pancreatic Surgery 77
7/14/2014 3:15:32 PM
the middle colic vein and SMV. The retroperitoneal attachments of the
small bowel and right colon mesentery are taken down to a much greater
extent in patients with uncinate tumors extending into the small bowel
mesentery. When necessary, the small bowel mesentery can be mobi-
lized by incising the visceral peritoneum all the way up to the ligament
of Treitz (Cattell-Braasch maneuver).
2. The Kocher maneuver is begun at the third part of the duodenum by
identifying the inferior vena cava. All tissue medial to the right gonadal
vein and anterior to the inferior vena cava is elevated along with the
pancreatic head and duodenum. This dissection is continued to the left
lateral edge of the aorta, with exposure of the anterior surface of the left
renal vein. A complete Kocher maneuver is necessary for the subsequent
dissection of the pancreatic head from the SMA (step 6). Particularly
important is the division of the leaf of peritoneum that extends from
the retroperitoneum to the root of mesentery; incision of this portion of
peritoneum is perhaps the most important part of the Kocher maneuver.
3. The portal dissection is commenced by exposing the CHA proximal and
distal to the right gastric artery and the gastroduodenal artery (GDA).
Both the right gastric and the GDA are then ligated and divided. Division
of the GDA allows mobilization of the hepatic (common-proper) artery
off the underlying PV. Cholecystectomy is then performed, and the com-
mon hepatic duct is transected at or above its junction with the cystic
duct. Following transection of the bile duct, bile cultures are sent and
indwelling endobiliary stents are removed. A bulldog clamp is placed on
the transected hepatic duct to prevent bile from soiling the right upper
quadrant until biliary reconstruction is completed. Note, the PV should
always be exposed and the location of the right hepatic artery noted prior
to dividing the common hepatic duct.
The PV should be identified but not extensively mobilized until step
6, at which time the stomach and pancreas have been divided. Care must
be taken to avoid injury to the superior pancreaticoduodenal vein drain-
ing the pancreatic head at the superolateral aspect of the PV, or significant
bleeding may occur when one does not yet have adequate exposure and
vascular control.
4. The terminal branches of the left gastric artery are ligated and divided
along the lesser curvature of the stomach prior to gastric transection. The
stomach is then transected with a linear gastrointestinal (GIA) stapler at
the level of the third or fourth transverse vein on the lesser curvature and
at the confluence of the gastroepiploic veins on the greater curvature to
complete a standard antrectomy (Fig. 7.1). The omentum is then divided
at the level of the greater curvature transection. In the pylorus-preserving
variant, the duodenum is divided just distal to the pylorus.
5. The loose attachments of the ligament of Treitz are taken down with care
to avoid injury to the inferior mesenteric vein (IMV) situated immedi-
ately to the patient’s left running caudal to cranial. The jejunum is then
transected with a linear GIA stapler approximately 8 to 10 cm distal to
the ligament of Treitz and its mesentery sequentially ligated and divided
with an energy device such as the LigaSure. This dissection is continued
proximally to involve the fourth and third portions of the duodenum. The
duodenal mesentery is divided to approximately the level of the aorta,
allowing the devascularized segment of duodenum and jejunum to be
reflected beneath the mesenteric vessels into the right upper quadrant.
6. This step is oncologically the most important and difficult part of the
operation. Traction sutures are placed on the superior and inferior
borders of the pancreas, and the pancreas is then transected with
002086376.INDD 78 7/14/2014 3:15:33 PM

FIGURE 7.1 Illustration demonstrating completion of step 3 and step 4. The porta hepatis
has been dissected, with ligation of the gastroduodenal and right gastric arteries. The
gallbladder has been removed and the common hepatic duct transected (step 3). The antrum
of the stomach has been divided at the level of the third or fourth transverse vein on the
lesser curvature (step 4).
e lectrocautery at the level of the PV. There is usually a small artery that
runs along the inferior border of the pancreas that is secured with the
traction sutures. If there is evidence of tumor adherence to the PV or SMV,
the pancreas is divided more upstream (to the tail) in the preparation for
segmental venous resection. The specimen is separated from the SMV
by ligation and division of the small venous tributaries to the uncinate
process and pancreatic head (Fig. 7.2). Complete removal of the uncinate
process from the SMV is required for full mobilization of the SMV–PV
confluence and subsequent identification of the SMA. Failure to fully
mobilize the SMV–PV confluence risks injury to the SMA and may result
in a positive SMA margin. In addition, without complete mobilization of
the SMV, it is difficult to expose the SMA. The inferior pancreaticoduode-
nal arteries (IPDAs) arising from the SMA must be identified and directly
ligated. Mass ligation of the IPDAs with mesenteric soft tissue is a com-
mon cause of postoperative hemorrhage as the vessels may retract with
the usual changes in blood pressure after extubation.
0002086376.INDD 79 7/14/2014 3:15:34 PM

FIGURE 7.2 Illustration of step 6. The pancreatic head and uncinate process are being
separated from the SMV–PV confluence. The pancreas has already been transected at
the level of the PV. Small venous tributaries from the PV and SMV are ligated and divided
including delicate branches from the uncinate to the first jejunal branch of the SMV.
Proper mobilization of the SMV involves identification of its first

jejunal branch. This branch originates from the right posterolateral aspect
of the SMV (at the level of the uncinate process), travels posterior to the
SMA, and enters the medial (proximal) aspect of the jejunal mesentery.
The jejunal branch may course anterior to the SMA in up to 20% of cases,
a situation that makes this dissection somewhat simpler. If tumor involve-
ment of the SMV (at the level of the jejunal branch) prevents dissection of
the uncinate process from the SMV, the jejunal branch should be divided
proximal to the site of tumor encasement and again at its junction with
the main trunk of the SMV. Once the uncinate is separated from the distal
SMV, medial retraction of the SMV–PV confluence allows exposure of the
SMA. The specimen is then separated from the right lateral wall of the SMA,
which is dissected to its origin at the aorta; the plane of dissection should
be directly on the adventitia of this vessel (Fig. 7.3).
The pancreatic and common hepatic duct transection margins are
submitted for frozen-section evaluation. Positive resection margins dem-
onstrating invasive carcinoma on the biliary or pancreatic duct mandate
002086376.INDD 80 7/14/2014 3:15:35 PM

FIGURE 7.3 Illustration demonstrating sharp dissection of the SMA margin, the most criti-
cal component of step 6. Medial retraction of the SMV–PV confluence facilitates dissection
of the soft tissues adjacent to the lateral wall of the proximal SMA. The IPDA (or arteries)
is identified at its origin from the SMA, ligated, and divided.
further resection until clear margins are achieved. Specimens should be

oriented for the pathologist. The SMA margin must be identified and inked
for the pathologist, as it cannot be accurately assessed in a retrospective
fashion.
Pancreatic, Biliary, and Gastrointestinal Reconstruction

Reconstruction after pancreaticoduodenectomy begins with the pancreatic
anastomosis.
1. The pancreatic remnant is mobilized from the retroperitoneum and
distal splenic vein for a distance of 2 to 2.5 cm to allow accurate suture
placement for the pancreaticojejunal anastomosis. The transected jeju-
num is brought retrocolic through a generous incision in the transverse
mesocolon to the left (rather than the right) of the middle colic vessels.
This minimizes tension on the anastomosis and provides a more direct
route for the small bowel mesentery to approximate the pancreatic rem-
nant than is afforded using a right-sided mesenteric window. A two-layer,
end-to-side, duct-to-mucosa pancreaticojejunostomy is performed.
0002086376.INDD 81 7/14/2014 3:15:36 PM

Occasionally, we use a one-layer anastomosis if the pancreas is small, is

firm, and contains a dilated duct. Additionally, a small silastic stent may
be used if the pancreatic duct is small. If the duct is too small to allow
passage of the stent, an invaginating technique dunking the pancreatic
remnant into the jejunal limb may rarely be required. The anastomosis
between the pancreatic duct and the small bowel mucosa is completed
with 5-0 monofilament absorbable sutures. Each stitch incorporates
a generous bite of the pancreatic duct and a full-thickness bite of the
jejunum. The posterior knots are tied on the inside and the lateral and
anterior knots on the outside. When a stent is used, this is placed into
the pancreatic duct and small bowel for a distance of approximately 2 to
3 cm prior to tying the anterior sutures.
2. A single-layer biliary anastomosis is performed using interrupted 4-0 or
5-0 absorbable monofilament sutures at a distance from the pancreati-
cojejunal anastomosis chosen to eliminate any tension. An interrupted
technique is utilized to avoid purse stringing of the anastomosis. A stent
is rarely ever used, even when the bile duct is of normal caliber.
3. An antecolic, end-to-side gastrojejunostomy is constructed in two layers.
A posterior row of 3-0 silk sutures is followed by a full-thickness inner
layer of running absorbable monofilament sutures; an anterior row of silk
sutures completes the anastomosis. The distance between the biliary and
gastric anastomoses should be at least 45 to 50 cm, allowing the jejunum
to assume its antecolic position for the gastrojejunostomy without ten-
sion and also minimizing the risk of reflux cholangitis. A 10-Fr feeding
jejunostomy tube may be placed approximately 30 cm distal to the gas-
trojejunostomy and used for feeding in the postoperative period.
Prior to closure, the abdomen is carefully irrigated with water, and the
small bowel is run to ensure proper alignment. The use of drains remains
controversial although many surgeons still drain both the hepaticojejunos-
tomy and the pancreaticojejunostomy. In patients undergoing neoadjuvant
therapy that includes radiation, the authors prefer to only drain the right
upper quadrant with a single silastic drain to capture lymphatic fluid in the
first 2 to 3 days after surgery. The falciform ligament, which was mobilized
at the start of the laparotomy, is now placed between the hepatic artery and
the afferent jejunal limb to cover the GDA stump. This maneuver minimizes
the risk of pseudoaneurysm formation at this site in the event of a pancreatic
anastomotic leak. A tongue of omentum may be used for the same purpose.
Pylorus Preservation
Pylorus preservation may be considered in patients with small periampullary
neoplasms but should not be performed in patients with large pancreatic
head tumors or in the setting of grossly positive pyloric or peripyloric lymph
nodes. The essential differences in technique (compared to standard pancre-
aticoduodenectomy) involve steps 3 and 4 described above. The duodenum is
divided approximately 2 to 3 cm beyond the pylorus with a linear GIA stapler
and the gastroepiploic arcade divided at that level. The staple line is removed
prior to creation of the duodenojejunostomy leaving approximately 2 cm of
the duodenum distal to the pylorus; we usually send the duodenal margin for
frozen section evaluation to exclude an unsuspected positive margin from
appearing as a surprise on the final pathology report. The pylorus is gently
dilated with a Kelly clamp or index finger (or both). The anastomosis is then
performed in an end-to-side fashion using a single- (author’s preference) or
double-layer technique with monofilament absorbable sutures. Placement of
a feeding jejunostomy in the setting of pylorus preservation may be consid-
ered due to the potentially increased incidence of delayed gastric emptying.
002086376.INDD 82 7/14/2014 3:15:36 PM

Venous Resection
Venous resection should only be performed in carefully selected patients
with tumor adherence to the SMV or SMV–PV confluence without evidence
of tumor encasement (>180 degrees) of the SMA or celiac axis. Steps 1
through 5 are completed as described above. Tumor adherence to the lateral
wall of the SMV–PV confluence prevents dissection of the SMV and PV off the
pancreatic head and uncinate process, thereby inhibiting medial retraction
of the SMV–PV confluence (and lateral retraction of the specimen). Division
of the splenic vein is performed when tumor abutment (on the lateral wall
opposite the splenic vein) or encasement is at the level of the splenic vein
junction. Division of the splenic vein allows complete exposure of the SMA
medial to the SMV and allows the retroperitoneal dissection to be completed
by sharp division of the soft tissues anterior to the aorta and to the right
of the exposed SMA. The specimen is then attached only by the SMV–PV
confluence. Vascular clamps are placed 2 to 3 cm proximal and distal to
the involved venous segment, and the vein is transected, allowing tumor
removal (Fig. 7.4). A 2- to 3-cm segment of SMV–PV confluence can safely be
resected without the need for interposition grafting as increased SMV and
PV length is provided when the splenic vein is divided. Venous resection is
always performed with inflow occlusion of the SMA and systemic heparin-
ization. It is important to mark the anterior surface of the SMV and PV prior
to venous resection to prevent an inadvertent anastomotic twist.
Upper gastrointestinal hemorrhage due to sinistral portal hyperten-
sion following splenic vein ligation can occur if the IMV enters the SMV
rather than the splenic vein—in which case there is inadequate venous out-
flow from the stomach and spleen due to splenic vein ligation. In contrast,
when the IMV enters the splenic vein, the IMV provides a route for collat-
eral venous flow from the ligated splenic vein in a retrograde fashion to the
systemic venous circulation. When the splenic vein must be divided and the
IMV enters the SMV, we create a distal splenorenal shunt to allow decom-
pression of the stomach and spleen into the systemic venous circulation.
Ideally, the splenic vein–PV junction should be preserved. However,
this is only possible when tumor invasion of the SMV or PV does not involve
the splenic venous confluence. Splenic vein preservation significantly limits
mobilization of the PV and prevents primary anastomosis after segmental
SMV resection unless the resection is limited to less than 2 cm. Therefore,
most patients who need SMV resection with splenic vein preservation require
an interposition graft. Our preferred conduit is the internal jugular vein
(IJV). Venous resection and reconstruction can be performed either before
the specimen has been separated from the right lateral wall of the SMA or
after complete mesenteric dissection. We first described the latter “artery-
first” approach in the manuscript by Leach et al., and this is our preferred
approach. Postoperatively, patients with vascular resection/reconstruction
are prescribed aspirin and subcutaneous heparin starting in the recovery
room assuming coagulation parameters are acceptable (INR < 1.5).
DISTAL PANCREATECTOMY
This operation can be performed laparoscopically, open, or with a hybrid
technique. A distal pancreatectomy can be completed from medial to lat-
eral (transect pancreas at PV–SMV–splenic vein confluence and dissection
completed from the patient’s right to left) or from lateral to medial (mobi-
lizing the distal pancreas and spleen off of the retroperitoneal attachments
and vasculature first, with pancreatic transection and dissection of the root
of small bowel mesentery as the final step). Regardless of which technique
is used, we always ligate the splenic artery before dividing the splenic vein.
0002086376.INDD 83 7/14/2014 3:15:36 PM

FIGURE 7.4 Illustration demonstrating resection of the SMV–PV confluence with splenic
vein preservation. Vascular clamps have been placed on the SMV below and the PV above
the involved venous segment, and a baby bulldog is utilized to occlude the splenic vein.
A Rummel tourniquet is placed on the SMA in preparation for excision of the SMV en bloc
with the specimen. The intact splenic vein tethers the PV, making a primary anastomosis
impossible in most cases. As such, the reconstruction will require a segmental vein graft
(we prefer a conduit of IJV that we prepare prior to the application of vascular clamps). In
order to reduce the possibility of performing the anastomosis with a twist in the conduit,
the IJV is marked while still in situ to ensure correct orientation.
Frequently, when dealing with a large tumor in the proximal body of the
pancreas, we may divide the pancreas before dividing the splenic artery—
and sometimes even before ligating the splenic artery. By dividing the
pancreas, exposure to the celiac artery and proximal splenic artery is sig-
nificantly improved.
Lateral to Medial Approach

The greater omentum is separated from the transverse colon, and the splenic
flexure is mobilized caudally. The peritoneum lateral to the spleen is then
incised in a cephalad direction up to the gastroesophageal junction (GEJ)
and the spleen and distal pancreas mobilized out of the left upper quad-
rant, as the spleen and pancreas are dissected free of their retroperitoneal
002086376.INDD 84 7/14/2014 3:15:37 PM

attachments to the kidney and adrenal gland. The short gastric vessels are
ligated and divided, allowing medial and cephalad retraction of the stom-
ach. Before division of the splenic vein, the splenic artery should be ligated,
even if it is not divided. The IMV may be divided if it enters the splenic vein,
or alternatively, the splenic vein can be divided just proximal (to the left) of
the IMV–splenic vein junction. The mesenteric root is divided from the left
lateral border of the SMV to the left lateral border of the aorta, staying ante-
rior to the SMA. Exposure of the SMA inferior to the neck of the pancreas
allows the dissection to proceed directly anterior to this vessel under direct
vision. If the splenic artery was not ligated earlier in the operation (author’s
preference), this is now completed. Again, the artery is divided before
ligation and division of the splenic vein. The pancreas is then transected
(before ligation of the splenic artery if increased exposure is needed), a step
that can be performed in a variety of manners. The most common tech-
niques include stapler transection or division with electrocautery followed
by suture ligation/closure. Regardless of technique, the most important
step is identification and suture ligation of the pancreatic duct (we prefer
5-0 monofilament suture). The staple line may also be reinforced with 5-0
monofilament sutures on pledgets. When the pancreas is divided further
medially than the level of the pancreatic neck, the point of pancreatic tran-
section is virtually always too thick for a stapling device.
After delivery of the specimen, we routinely bring the mobilized falci-
form ligament through the lesser omentum to cover the stump of the splenic
artery and the pancreatic transection site—analogous to the use of the fal-
ciform ligament following pancreaticoduodenectomy discussed previously.
Medial to Lateral Approach

After entering the lesser sac through the gastrocolic ligament, the inferior
border of the pancreas is identified, and the peritoneum is incised. The dis-
section is carried medially until the SMV is identified as it passes beneath
the pancreas. A tunnel along the anterior surface of the SMV–PV confluence
can then be created, and the superior border of the pancreas is freed from
the hepatic artery and PV. During laparoscopy, passing a sling underneath
the pancreas and elevating the gland can aid in the further exposure of the
pancreatic body and tail as the dissection proceeds laterally. Once sufficient
space has been created for the passage of a stapling device, the pancreas is
transected. It is then reflected to the left, facilitating its complete dissection
from its retroperitoneal attachments.
Splenic Preservation
Distal pancreatectomy may also be performed with preservation of the
spleen. Two techniques are available in this context, namely, preservation
of the spleen via retrograde flow through the short gastric vessels (Warshaw
technique) or by preserving the entire length of the splenic artery and vein.
The Warshaw technique involves entering the lesser sac through the gas-
trocolic omentum but halting the dissection prior to encountering the left
gastroepiploic and short gastric vessels. The avascular plane behind the
pancreas is then mobilized starting at the inferior border of the gland near
its tail, anterior to the splenic vessels. Dissection continues until the supe-
rior border is freed from the retroperitoneum. The tail of the pancreas can
then be gently reflected to the patient’s right, away from the splenic hilum,
exposing the vessels for stapler fixation. After division, the dissection of the
pancreas (and splenic vessels) proceeds to the right as far as is required
to clear the indicated pathology. The pancreatic substance is then divided
with a stapler (including the splenic vessels en masse), allowing delivery of
the specimen.
0002086376.INDD 85 7/14/2014 3:15:37 PM

Preservation of the splenic vessels is only applicable for certain tumor

histologies (i.e., mucinous cystic neoplasms that lack malignant character-
istics or symptomatic serous cystadenomas), as dissection in this plane vio-
lates oncologic principles of en bloc resection of regional lymphatic tissue.
This procedure involves meticulous dissection of the multiple small vessels
away from the pancreatic parenchyma that arise from the splenic artery or
drain to the splenic vein. This is usually performed from medial to lateral to
avoid the labyrinth of tiny vessels often present at the splenic hilum.
Laparoscopic Approach
Any of the approaches described above for distal pancreatectomy may be
performed laparoscopically. A Hasson cannula placed supraumbilically
allows visualization with a 30-degree laparoscope. A 5-mm working port
is placed in the right upper quadrant approximately in the mid-clavicular
line, and a second port of 12 mm (to allow passage of an Endo GIA stapler)
is placed to the left of the midline at approximately the same level. Another
5-mm port for retraction is often placed further laterally on the left side
for the surgical assistant. An umbilical tape or vessel loop passed under-
neath the pancreas aides in retraction and dissection of the specimen. The
resection is then completed as described above with pancreatic transection
completed using a stapler with seam guard (Gore).
Hybrid Approach
A laparoscopic-assisted hybrid technique combines the benefits of mini-
mally invasive surgery with the security of precise and direct ligation of the
pancreatic duct and splenic artery. The patient is positioned in the right lat-
eral decubitus position (left side up). A Hasson cannula is placed at the umbi-
licus, pneumoperitoneum is achieved, and a standard four-port laparoscopy
is undertaken. The splenic flexure of the colon is mobilized and the greater
omentum dissected from the distal transverse colon. The peritoneum lateral
to the spleen is incised proceeding cephalad until reaching the GEJ from the
patient’s left side. The dissection then continues lateral to medial mobiliz-
ing the spleen and tail of the pancreas toward the midline. The left crus of
the diaphragm marks the medial most extent of the laparoscopic dissec-
tion. The short gastric arteries are divided. The patient is then returned to
the supine position and a limited upper midline incision created, and the
operation proceeds as for the medial to lateral approach. The splenic artery
is identified at the superior border of the pancreas and ligated (either before
or after pancreatic transection). The splenic vein–SMV junction is identi-
fied at the inferior border of the pancreatic body and, after dissection for a
short distance, divided with an Endo GIA stapling device either flush with
the SMV–PV confluence or just proximal to the IMV–splenic vein junction.
Mobilization of the inferior aspect of the pancreas is now completed and the
specimen freed from the retroperitoneum. After selection of an appropriate
point of transection, the pancreas is divided as described above.
TOTAL PANCREATECTOMY
Total pancreatectomy is accomplished by following steps 1 and 3 of our six-
step pancreaticoduodenectomy. After identification of the infrapancreatic
SMV and completion of the portal dissection, the proximal splenic artery
is exposed and ligated. Resection as described for distal pancreatectomy is
then completed. While dissection of the SMA occurs from both the patient’s
left and right side, it is important, when possible, to leave some of the auto-
nomic nerves on the adventitia of the SMA to preserve small bowel inner-
vation and prevent rapid gastrointestinal transit. Division of the splenic
002086376.INDD 86 7/14/2014 3:15:37 PM

vein is often delayed until the completion of steps 4 (gastric or duodenal

transection) and 5 (ligament of Treitz, proximal jejunum, and duodenum).
Then, in preparation for completion of the retroperitoneal dissection, the
spleen and distal pancreas are again reflected medially, exposing the ori-
gin of the splenic vein–SMV junction; the splenic vein is then ligated and
divided. Step 6 of pancreaticoduodenectomy is then completed in the stan-
dard manner and the specimen delivered intact. With the entire pancreas
removed, reconstruction is a simpler procedure, involving only biliary and
gastric anastomoses.
CENTRAL PANCREATECTOMY
In rare patients with relatively small tumors of the pancreatic neck of favor-
able histology (i.e., benign-appearing mucinous cystic neoplasms, branch
duct IPMNs, solid pseudopapillary neoplasms), segmental resection of the
pancreatic neck and proximal body with preservation of the splenic artery
and vein may be utilized to preserve islet cell function. The proximal pan-
creas and pancreatic duct are oversewn as described above for distal pan-
createctomy. Reconstruction includes Roux-en-Y pancreaticojejunostomy
to the remaining segment of the distal pancreas. The small bowel is divided
35 to 45 cm distal to the ligament of Treitz. The distal limb is brought
through the transverse colon mesentery (retrocolic) and sewn to the dis-
tal pancreas as a two-layer, end-to-side, duct-to-mucosa pancreaticoje-
junostomy as described for pancreaticoduodenectomy. Additionally, the
oversewn distal end of the pancreatic head is reinforced by buttressing the
serosa of Roux limb, distal to the pancreaticojejunostomy, to the pancreatic
transection site. The proximal pancreaticobiliary limb is connected down-
stream to restore intestinal continuity as a side-to-side jejunojejunostomy.
SUMMARY
A preoperative evaluation that includes detailed patient assessment for
comorbid conditions, functional status, and tumor marker trends and
assessment of resectability by objective CT criteria minimizes rates of non-
therapeutic laparotomy in patients with pancreatic cancer. The operative
techniques detailed in this chapter are utilized to minimize intraopera-
tive and perioperative complications. Pancreaticoduodenectomy and total
pancreatectomy alone or in combination with vascular resection are tech-
nically demanding and should not be undertaken without thorough under-
standing of the nuances of upper abdominal anatomy. Best outcomes are
obtained in high-volume centers by multidisciplinary teams that include
experienced pancreatic surgeons, anesthesiologists, and nurses attuned
to the intricacies of operative technique. Utilization of postoperative care
pathways promotes early recognition of complications. As improved sys-
temic therapies increase survival duration for patients with localized pan-
creatic cancer, the management of local-regional disease with surgery will
become even more complex and the role for extended resection to include
vascular resection and reconstruction considered with increased frequency.
Patient selection and detailed preoperative planning are simply invaluable
to insure the desired results.
Suggested Readings
Balachandran A, Darden DL, et al. Arterial variants in pancreatic adenocarcinoma.
Abdom Imaging 2008;33(2):214–221.
Christians KK, Lal A, et al. Portal vein resection. Surg Clin North Am 2010;90(2):309–322.
Christians KK, Tsai S, Tolat PP, et al. Critical steps for pancreaticoduodenectomy in
the setting of pancreatic adenocarcinoma. J Surg Oncol 2013;107(1):33–38.
0002086376.INDD 87 7/14/2014 3:15:38 PM

Evans DB, Farnell MB, et al. Surgical treatment of resectable and borderline
resectable pancreas cancer: expert consensus statement. Ann Surg Oncol
2009;16(7):1736–1744.
Katz MH, Hwang R, et al. Tumor-node-metastasis staging of pancreatic adenocarci-
noma. CA Cancer J Clin 2008;58(2):111–125.
Katz MH, Fleming JB, et al. Anatomy of the superior mesenteric vein with special ref-
erence to the surgical management of first-order branch involvement at pancre-
aticoduodenectomy. Ann Surg 2008;248(6):1098–1102.
Katz MH, Pisters PW, et al. Borderline resectable pancreatic cancer: the importance
of this emerging stage of disease. J Am Coll Surg 2008;206(5):833–846; discussion
846–838.
Katz MH, Varadhachary GR, et al. Serum CA 19-9 as a marker of resectability and
survival in patients with potentially resectable pancreatic cancer treated with
neoadjuvant chemoradiation. Ann Surg Oncol 2010;17(7):1794–1801.
Leach SD, Davidson BS, Ames FC, et al. Alternative method for exposure of the ret-
roperitoneal mesenteric vasculature during total pancreatectomy. J Surg Oncol
1996;61(2):163–165.
002086376.INDD 88 7/14/2014 3:15:38 PM

8 Minimally Invasive
Pancreatic Surgery
Michael L. Kendrick
INTRODUCTION
Minimally invasive pancreatectomy is rapidly gaining favor in centers
specializing in pancreatic surgery. A growing body of evidence suggests sev-
eral outcome advantages over open approaches; these findings continue to
engender enthusiasm for laparoscopy. Laparoscopic distal pancreatectomy
(LDP) is the most widely adopted procedure, as it requires basic laparo-
scopic skills of dissection and no reconstruction. Procedures with more
complex resection or reconstruction have seen more gradual acceptance.
The predominant challenge preventing the wide application of minimally
invasive surgery (MIS) approaches in pancreatic surgery continues to be
the lack of adequate surgeon training. Few centers have sufficient experi-
ence with more advanced procedures such as pancreaticoduodenectomy
and total pancreatectomy; thus, large comparative trials of laparoscopy
versus open approach are lacking for these procedures. However, the
increasing evidence in support of MIS approaches for all pancreatic opera-
tions continues to emerge as larger series are reported, and more centers
develop minimally invasive programs.
INDICATIONS AND CONTRAINDICATIONS

The indications for MIS pancreatectomy mirror those for open approaches
and include all benign and malignant conditions requiring pancreatic
resection. Expanding the indications for pancreatic resection solely
based on the ability to perform it with minimally invasive approaches is
not recommended. Absolute contraindications for MIS pancreatectomy
are also similar to open approaches and include patients with prohibitive
comorbidities or poor functional status. The majority of reported contra-
indications are therefore relative, and experience of the surgeon is the key
determinate (Table 8.1).
TECHNIQUE
Laparoscopic Distal Pancreatectomy
Setup, Access, and Initial Exposure
The patient is positioned supine with 15 degrees of reverse Trendelenburg.
Whereas some advocate a right lateral decubitus position or place a bump
under the left flank, we prefer a supine position that facilitates a right-to-
left dissection. A total of four trocars are placed as depicted in Figure 8.1.
A periumbilical incision is made, the trocar placed under direct vision, and
a pneumoperitoneum established. The remaining three trocars are placed
under laparoscopic visualization. The gastrocolic ligament is divided
peripheral to the epiploic vessels using the harmonic scalpel, and the lesser
sac is entered. The short gastric vessels are preserved, and the splenic flex-
ure of the colon is mobilized inferiorly. Visual inspection of the anterior
89
0002152322.INDD 89 7/14/2014 3:19:24 PM

TABLE
8.1
Absolute
Contraindications for Laparoscopic Pancreatic Resection
Relative
Prohibitive medical comorbidities Locally advanced malignancy
Poor functional status Proximity to major vasculature
Severe obesity
Prior major abdominal operations
aspect of the pancreas is performed, and the lesion is identified either

through palpation or through the use of laparoscopic ultrasonography. The
stomach is sutured anteriorly to the retrocostal peritoneum to facilitate
exposure.
Pancreatic Mobilization
The transverse mesocolon is dissected off the inferior border of the pan-
creas, and the pancreas is dissected out of the retroperitoneum in the antic-
ipated area of transection. Early focus of dissection at the site of anticipated
transection and splenic vessels allows excellent access and exposure for a
safe dissection. In the event of bleeding during splenic vessel preservation,
the vessels can be quickly ligated, reducing blood loss. The splenic vein is
dissected off the posterior aspect of the pancreas. The splenic artery is iden-
tified and dissected either through an anterior approach at the superior
border of the pancreas or via a posterior-inferior approach after elevating
5 mm
5 mm 5 mm
12 mm
12 mm
FIGURE 8.1 Trocar positioning for LDP.
002152322.INDD 90 7/14/2014 3:19:26 PM

Chapter 8 / Minimally Invasive Pancreatic Surgery 91
the pancreas out of the retroperitoneum. The latter approach is most com-
monly used when the parenchymal transection is anticipated to occur
within the pancreatic body rather than the tail.
Pancreatic Resection
With the pancreas elevated off of the splenic vessels, the pancreatic
parenchyma is then divided with either a linear stapler with biologic rein-
forcement or the harmonic scalpel. If spleen preservation is planned, the
assistant retracts the pancreas anteriorly and laterally to expose the dis-
section plane between the pancreas and splenic vessels. Most tributary ves-
sels can be divided with the harmonic scalpel. Larger vessels (≥2 mm) are
ligated with suture or are clipped and divided. When planning splenectomy,
the splenic vessels are ligated and divided near the pancreatic transection
site prior to a right-to-left dissection with division of the short gastric ves-
sels and splenic peritoneal attachments.
Specimen Retrieval
The specimen is placed in an endobag and removed via the periumbilical
incision that is extended just long enough to accommodate the specimen.
The specimen is inspected on the back table, and a separate pancreatic mar-
gin is harvested. We routinely use frozen section histology for evaluation of
the primary lesion and the margin prior to termination of the procedure.
The extraction site is closed with interrupted suture, leaving the cephalad
two sutures untied. The trocar is reintroduced and the carbon dioxide (CO2)
pneumoperitoneum reestablished.
Pancreatic Stump Management and Closure

When the transection is not performed with the stapler, our preferred
method of pancreatic stump treatment is with saline-coupled radiofre-
quency energy source (Salient Surgical Technologies, Portsmouth, NH). The
pancreatic resection bed and exposed splenic vessels are then reinspected,
and hemostasis is ensured. Prophylactic operative drains are not routinely
used in the author’s practice. The trocars are removed, and sites are closed
with a 4-0 subcuticular monofilament absorbable suture.
Laparoscopic Pancreaticoduodenectomy
Setup, Access, and Initial Exposure
The patient is positioned supine in 15 degrees of reverse Trendelenburg.
Initial access is gained through a left subcostal site using a 12-mm transpar-
ent, cone-tip trocar. A CO2 pneumoperitoneum to 15 mm Hg is established,
and all visible peritoneal and visceral surfaces are inspected. Five additional
12-mm trocars are placed (Fig. 8.2).
The gastrocolic ligament is divided with the harmonic scalpel (Ethicon
Endosurgery, Cincinnati, OH), widely exposing the lesser sac. The gastroepi-
ploic vessels are clipped and divided with the harmonic scalpel. The duode-
nocolic ligament is resected en bloc with the specimen to facilitate regional
lymphadenectomy. A fan retractor is placed under the gastric antrum to
provide exposure of the pancreatic head and neck.
Pancreatic Neck Dissection

The pancreatic neck is approached with the surgeon positioned on the left
side of the patient. The common hepatic artery lymph nodes are removed,
and the common and proper hepatic arteries are identified to verify tumor
clearance. The portal vein is exposed at the cephalad border of the pan-
creatic neck and the superior mesenteric vein at the caudal border. The
0002152322.INDD 91 7/14/2014 3:19:26 PM

10 mm
10 mm
10 mm
10 mm 10 mm
10 mm
5 cm
FIGURE 8.2 Trocar positioning for laparoscopic

pancreaticoduodenectomy.
g astroduodenal artery is dissected, ligated, clipped, and divided. A plane

between the posterior aspect of the pancreatic neck and the portal vein/
superior mesenteric vein is developed. An articulating grasper is passed
posteriorly and an umbilical tape is placed around the pancreatic neck and
secured.
Duodenal Mobilization
The first portion of the duodenum is cleared, and the right gastric artery
is ligated and divided. The duodenum is divided 2 cm distal to the pylorus
using a linear stapler. The transverse colon is reflected cephalad, and the
area of the ligament of Treitz is dissected mobilizing the third and fourth
portions of the duodenum off of the aorta and inferior vena cava. The jeju-
num is divided 15 cm distal to the ligament of Treitz with the linear sta-
pler, and the jejunal mesentery is then divided with the harmonic scalpel
back to the uncinate process. The hepatic flexure of the colon is mobilized
inferiorly, and the ascending and transverse colon are retracted inferiorly
with a fan retractor. The surgeon moves to the patient’s right side, and
the Kocherization of the duodenum is extended cephalad to the hepatic
hilum.
Hepatic Hilar Dissection

A cholecystectomy is performed, leaving the cystic duct intact. The hilum
of the liver is dissected, and the common hepatic duct is identified, ligated
distally, and divided. All hilar lymphatic tissue is dissected inferiorly with
the specimen. The lateral aspect of the portal vein is cleared caudally until
002152322.INDD 92 7/14/2014 3:19:27 PM

the superior pancreaticoduodenal vein is identified, ligated, and divided.

The jejunal stump is brought into the supramesocolic compartment, and
the anterolateral aspect of the superior mesenteric vein is dissected proxi-
mally to the inferior border of the pancreas.
Pancreatic Head Resection

The pancreatic neck is divided with the harmonic scalpel, with the excep-
tion of the pancreatic duct, which is divided sharply with scissors. Using
the umbilical tape for retraction, the pancreatic head and uncinate process
are dissected off the portal vein, superior mesenteric vein, and superior
mesenteric artery (SMA). The inferior pancreaticoduodenal artery and vein
are ligated or clipped and divided with harmonic scalpel. Dissection of the
uncinate is performed adjacent to the adventitia of the SMA to assure an
appropriate oncologic margin.
Specimen Removal
The specimen is placed into an endobag and removed via the infraumbili-
cal trocar site, which is typically extended to a total length of 3 to 5 cm to
accommodate the specimen. Specimen inspection is performed on the
back table; separate pancreatic neck and bile duct margin are harvested
and sent for frozen section analysis. The portal vein groove and SMA mar-
gins are inked. The extraction site is closed with interrupted suture and a
CO2 pneumoperitoneum reestablished.
Anastomotic Reconstruction
The jejunum is brought through the duodenal resection bed (retromes-
enteric tunnel). An end-to-side pancreaticojejunostomy, duct-to-mucosa
anastomosis, is constructed with an inner layer of interrupted 5-0 Vicryl
suture and an outer layer of interrupted 3-0 PDS suture.
Approximately 10 cm distal to the pancreaticojejunostomy, the hepat-
icojejunostomy is performed. The surgeon stands on the patient’s right with
instruments and camera in the right-most trocars. This setup allows sewing
toward the surgeon as would be done in an open approach. The end-to-side
hepaticojejunostomy is constructed with a single layer of interrupted (duct
size ≤6 mm) or running (duct size >6 mm) 5-0 Vicryl suture.
An antecolic end-to-side duodenojejunostomy is constructed approxi-
mately 40 cm distal to the hepaticojejunostomy, using two layers of running
3-0 Vicryl suture.
A single, 5-mm round, closed-suction drain is brought through the
right abdominal wall and positioned posterior to the hepaticojejunostomy
and anterior to the pancreaticojejunostomy. The trocars are removed under
direct vision and the skin incisions closed with a subcuticular 4-0 monofila-
ment absorbable suture.
POSTOPERATIVE MANAGEMENT
Postoperative care after MIS pancreatectomy is similar for both distal
pancreatectomy and pancreaticoduodenectomy. The orogastric tube is
removed at the end of the procedure. Patients are started on clear liquids
on the first postoperative day, and diet is advanced over the next 48 hours.
Pain is controlled with intravenous ketorolac scheduled every 6 hours and
with patient-controlled intravenous administration of morphine. The drain
is removed on postoperative day 4 if drain amylase is low and no other
signs of pancreatic fistula are present. Hospital discharge is allowed on or
after the 5th postoperative day if the patient is tolerating a soft diet and is
without evidence of complication.
0002152322.INDD 93 7/14/2014 3:19:27 PM

OUTCOMES
Several large series of minimally invasive distal pancreatectomy have dem-
onstrated the feasibility, safety, and favorable outcomes of this approach.
Adequately powered, randomized controlled trials of LDP versus open distal
pancreatectomy (ODP) have not been performed, and the assessment of the
outcomes of these two approaches is predominantly limited to retrospective
comparative trials. Recently, a large meta-analysis of 18 studies including
1,814 patients suggested that minimally invasive distal pancreatectomy is
associated with less blood loss, reduced overall and wound-specific compli-
cation rates, and shorter length of hospital stay. Importantly, no differences
in operative time, margin status, pancreatic fistula rates, or mortality were
seen between the two groups.
To further elucidate whether risk factors for operative morbidity dif-
fer between LDP and ODP, Cho and colleagues performed an analysis of
data from nine separate academic centers comparing LDP and ODP. Of
693 patients undergoing distal pancreatectomy (LDP = 254, ODP = 439),
multivariate analysis demonstrated that BMI ≤ 27, nonadenocarcinoma,
and pancreatic specimen length ≤8.5 cm had higher rates of fistula after
ODP than after LDP. This study identified no preoperatively variables asso-
ciated with the increased risk of pancreatic fistula after LDP compared to
ODP. Additionally, no patient cohorts were identified that had higher rates
of postoperative complication for LDP than for ODP.
Laparoscopic Pancreaticoduodenectomy
Gagner and Pomp published the first report of laparoscopic pancreaticoduo-
denectomy in 1994 and subsequently published a series of 10 patients in 1997.
The conversion rate for this series was 40% and the operative time was long
(8.5 hours). The authors concluded that there were no perceivable advantages
of the MIS approach. Over a decade passed before the first substantial series
of total laparoscopic pancreaticoduodenectomy (TLPD) was published by
Palanivelu and colleagues. Of 42 patients undergoing TLPD, the indication
was malignancy in 95%. With a mean operative time of 370 minutes, estimated
blood loss of 65 mL, and pancreatic fistula rate of 7%, this report not only
established the feasibility but also clearly set a challenge to investigate this
technique further. To date, only six published series have reported more than
50 patients undergoing TLPD and are shown in Table 8.2. While the outcomes
appear comparable to those reported for open approaches, appropriate and
intentional selection bias in these early experiences limits validation of equiv-
alency or potential advantages. Asbun and colleagues recently reported a
retrospective, comparative analysis of laparoscopic and open pancreaticodu-
odenectomy. Patients undergoing the laparoscopic approach had significantly
less estimated blood loss, fewer transfusions, and shorter length of hospital
stay. No differences in overall or pancreas-specific complications were seen;
however, operative time was longer for the laparoscopic group. Our institu-
tion has now performed laparoscopic pancreaticoduodenectomy in over 300
patients, and a comparative analysis is forthcoming. We continue to observe
the typical advantages of MIS approaches with TLPD as reported earlier.
Further investigation should be in evaluating the potential impact of laparo-
scopic approaches on the quality-of-life, oncologic, and long-term outcomes.
Robotic-Assisted Laparoscopic Pancreatic Resection

Robotic assistance has been described as a useful adjunct to pure lapa-
roscopic approaches and is currently gaining interest. Several series have
published outcomes of robotic-assisted pancreatic resection. Zureikat and
002152322.INDD 94 7/14/2014 3:19:27 PM

0002152322.INDD 95
TABLE
Selected Series of Laparoscopic Pancreaticoduodenectomya
8.2
Publication Year No. of Patients Procedure Op Time (Min) EBL (mL) PF (%) DGE (%) LOS (Days) Mortality (%)
Palanivelu 2009 75 TLPD 357 74 7 – 8 1.3
Kendrick 2010 62 TLPD 368 240 18 15 7 1.6
Giulianotti 2010 60 RALPD 421 394 21 5 22 3
Kim 2012 100 TLPD 487 – 6 2 20 1
Asbun 2012 53 TLPD 541 195 10 11 8 6
Zureikat 2013 132 RALPD 527 300 7 – 10 1.5
Totalb 482 453 263 14 9 13 2
a
Includes only series with ≥50 patients and data sufficient for analysis.
b
Weighted averages.
TLPD, total laparoscopic pancreaticoduodenectomy; RALPD, robotic-assisted laparoscopic pancreaticoduodenectomy; EBL, estimated blood loss; PF, pancreatic fistula; DGE, delayed gastric emptying;
LOS, length of hospital stay.
7/14/2014 3:19:28 PM
colleagues reported a large single institutional series of robotic pancreatec-

tomy. Of 250 consecutive robotic pancreatic resections, the most common
operations included are pancreaticoduodenectomy (n = 132), distal pancre-
atectomy (n = 83), central pancreatectomy (n = 13), enucleation (n = 10), and
total pancreatectomy. In these selected patients, the safety and feasibility
were established with comparable morbidity and mortality rates compared
to those reported for open approaches.
In a comparative study of LDP versus robotic-assisted LDP, Kang and
colleagues reported greater success with spleen preservation with the use
of robotic assistance (56% vs. 5%, P = 0.027). Significant differences in the
study group included a younger age and disease indication. Operative time
(349 vs. 258 minutes, P = 0.016) was longer, and cost (USD 8,300 vs. 3,862)
was significantly higher for the robotic group compared to the laparoscopic
group. The small sample size and patient selection may account for these
findings. Patients in the robotic-assisted group were also younger and less
likely to have intraductal papillary mucinous neoplasm, where pericystic
inflammation is frequently noted.
Reported advantages and disadvantages of the robotic platform are
listed in Table 8.3. Pertinent advantages may include the relative ease of
training and potential avoidance of prerequisite advanced laparoscopic
skills. The purported advantages of a three-dimensional view, added maneu-
verability, and fine motor movement, while enticing, are immeasurable with
regard to their independent effect on outcomes over pure laparoscopic
approaches. Significant disadvantages of the current robotic systems are the
lack of haptic feedback and the cost of the equipment and its maintenance.
With the feasibility and favorable outcomes of pure laparoscopic
approaches having already been established, additional comparative trials
of robotic compared to open approaches are of questionable value. The per-
tinent questions are whether the addition of robotic assistance improves
outcomes over pure laparoscopic approaches and does the increase cost
justify its use? The current literature lacks any adequately powered, appro-
priately designed trial to address these important questions. Further, the
answers are unlikely to be forthcoming for the most complex procedures
such as pancreaticoduodenectomy due to the fact that few centers will be
able to perform both approaches in sufficient numbers to allow a meaning-
ful comparison.
At present, the main advantage of robotic assistance appears to be cen-
tered toward the surgeon rather than the patient. Clearly, surgeons without
sufficient laparoscopic skills will be more able to perform complex laparo-
scopic procedures using robotic assistance. A more widespread application
of MIS approaches will provide advantages for patients. Another potential
advantage requiring further investigation is the ergonomics of surgeon posi-
tioning, which may reduce surgeon fatigue and chronic injuries c ompared
to pure laparoscopic approaches.
TABLE
8.3
Advantages
Advantages and Disadvantages of Robotic-Assisted Approaches
Disadvantages
“Intuitive”—replicates open skills Lack of haptic feedback
Increased range of motion Expense (purchase, maintenance)
Fine motor movement Set-up time
Three-dimensional view Surgeon remote from the patient
Surgeon ergonomics Loss of bedside “peripheral” view
002152322.INDD 96 7/14/2014 3:19:28 PM

Laparoscopic Approaches to Pancreatic Malignancy

A valid concern among many skeptics of minimally invasive approaches to
pancreatic resection is the ability of these approaches to maintain onco-
logic principles. Given the relative recent application of MIS approaches in
pancreatic malignancy, long-term oncologic outcomes are not available.
Surrogates to assess the quality of the oncologic resection include margin
status, number of lymph nodes resected, and short-term recurrence and
survival.
In a multicenter, retrospective, matched cohort analysis of patients
undergoing laparoscopic versus ODP for adenocarcinoma, Kooby and col-
leagues identified no difference in oncologic outcomes based on approach.
Specifically, margin-negative status (74% vs. 66%, P = 0.61), number of lymph
nodes harvested (14 vs. 12, P = 0.41), and median survival (16 vs. 16 months,
P = 0.71) were similar.
We have recently evaluated the outcomes of open and laparoscopic
pancreaticoduodenectomy at our institution from 2008 through 2013. To
avoid the confounding factors of invasiveness, margin-negative rates, and
survival among various types of malignancy, we evaluated only patients
with pancreatic ductal adenocarcinoma. Of 322 patients undergoing pan-
creaticoduodenectomy specifically for pancreatic adenocarcinoma, 108
underwent TLPD and 214 underwent OPD. Operative time (379 vs. 387 min-
utes) and need for major venous resection (13% vs. 22%) were not different.
The estimated blood loss and length of hospital stay were less for LPD com-
pared to OPD (492 vs. 867 mL, P = <0.001, and 6 vs. 9 days, P = <0.001, respec-
tively). With regard to oncologic outcomes, tumor size, margin-negative
resection, and number of lymph nodes harvested were not different.
A significantly smaller proportion of patients in the TLPD group had
a delay greater than 8 weeks from operation to chemotherapy. Despite
similar rates of margin-negative resection, local disease as the initial site
of recurrence was less common in the TLPD group (15% vs. 27%). While
there was no difference in overall survival, TLPD patients had a longer
progression-free survival.
Major vascular resection during pancreaticoduodenectomy for pancre-
atic head adenocarcinoma has been described as feasible and safe in open
approaches. Primary objectives of major venous resection are to increase
the number of patients eligible for a potentially curative procedure and to
increase the rate of R0 resections. Until very recently, the need for vascu-
lar resection had been a contraindication for laparoscopic approaches. Our
institution reported an early experience suggesting the feasibility and safety
of laparoscopic major venous resection at the time of pancreaticoduode-
nectomy. More recently, a comparative analysis of laparoscopic and open
pancreaticoduodenectomy with major venous resection demonstrated less
blood loss with comparable graft patency and complication rates for the
laparoscopic approach.
CONCLUSION
Minimally invasive approaches for pancreatic resection are feasible and
safe. Whereas level I evidence is lacking, existing comparative trials suggest
noninferiority and possible advantages for laparoscopic compared to open
approaches for both benign and malignant disease. Consistently reported
advantages include reduced blood loss and shorter length of hospital stay
for most procedures. A more widespread acquisition of advanced laparo-
scopic skills or access to robotic platforms will continue to advance the use
of minimally invasive approaches for pancreatic surgery. Oncologic and
quality of life outcomes are needed to assess value in addition to the typical
0002152322.INDD 97 7/14/2014 3:19:28 PM

advantages reported for MIS approaches. Careful scrutiny should continue

to assure appropriate outcomes and early identification of unexpected
complications and to substantiate the perceived advantages of minimally
invasive approaches for pancreatic resection.
Suggested Readings
Asbun HJ, Stauffer JA. Laparoscopic vs. open pancreaticoduodenectomy: overall
outcomes and severity of complications using the Accordion Severity Grading
System. J Am Coll Surg 2012;215:810.
Kendrick ML. Laparoscopic and robotic resection for pancreatic cancer. Cancer J
2012;18:571.
Kendrick ML, Sclabus GM. Major venous resection during total laparoscopic pancre-
aticoduodenectomy. HPB (Oxford) 2011;13:454.
Kooby DA, Hawkins WG, Schmidt CM, et al. A multicenter analysis of distal pancre-
atectomy for adenocarcinoma: is laparoscopic resection appropriate. J Am Coll
Surg 2010;210:779.
Venkat R, Edil BH, Schulick RD, et al. Laparoscopic distal pancreatectomy is associ-
ated with less overall morbidity compared to the open technique: a systematic
review and meta-analysis. Ann Surg 2012;255:1048.
Zureikat AH, Moser AJ, Boone BA, et al. 250 robotic pancreatic resections: safety and
feasibility. Ann Surg 2013;258:554.
002152322.INDD 98 7/14/2014 3:19:28 PM

9 Complications of Pancreatic
Surgery
Timothy R. Donahue and Howard A. Reber
INTRODUCTION
At high-volume centers, the operative mortality after pancreaticoduode-
nectomy (PD) is less than 3%. However, morbidity remains high with an
overall rate of postoperative complications ranging from 30% to 65%. These
complications can lead to prolonged hospital stays, increased readmission
rates, and greater hospital costs. Moreover, in patients who undergo an
operation for pancreatic cancer, adjuvant chemotherapy is delayed. Thus, it
is important to minimize the complication rate and, when a complication
occurs, to treat it promptly and effectively. The complications of pancreatic
surgery include pancreatic fistula, delayed gastric emptying (DGE), hemor-
rhage, biliary fistula, and pancreatic exocrine and endocrine insufficiency.
In this section, we will review the diagnosis and treatment of each of these
complications.
PANCREATIC FISTULA
Definition and Incidence
A pancreatic fistula is an abnormal communication between the pancreas
and adjacent or distant organs or spaces (internal fistula) or the skin (exter-
nal fistula). Fistulas comprise amylase-rich exocrine pancreatic secretions,
and external fistulas are the most common cause of prolonged morbidity
and mortality associated with pancreatic surgery. After PD, a fistula forms
due to impaired healing of the pancreatic anastomosis or, after distal pan-
createctomy, incomplete healing of the cut edge of the pancreas. Fistulas
can also form after middle (segmental) pancreatectomy or enucleations of
tumors, particularly in the latter case if the main pancreatic duct is injured.
Fistulas are arbitrarily described as low output if the volume is less than
200 mL/d and high output if it is greater than 200 mL/d.
Because of variability in how pancreatic fistula has been defined,
there is a wide range of the rate of pancreatic fistula formation reported in
the literature. To standardize these discrepancies, the International Study
Group on Pancreatic Fistula (ISGPF) published a consensus definition in
2005. Postoperative pancreatic fistula was defined as drain output of any
measurable volume after postoperative day 3 with an amylase content of
at least three times the upper limit of normal in the serum. The ISGPF also
developed a grading system (A, B, or C) based on the clinical impact and
additional complications caused by the fistula (Table 9.1). Grade A fistulas
pose little or no burden on the patient, while grade C fistulas are associated
with significant morbidity and occasional mortality.
Numerous risk factors contribute to fistula development. The most
widely accepted risks include a soft pancreatic texture that does not hold
sutures well for the pancreatic anastomosis in PD or sutures or staples for
remnant closure in distal pancreatectomy, a small pancreatic duct (<3 mm
99
0002086377.INDD 99 7/14/2014 3:36:03 PM

TABLE
Main Parameters for Postoperative Pancreatic Fistula
9.1 Grading
Grade A B C
Clinical conditions Well Often well Ill appearing/bad
Specific treatmenta No Yes/no Yes
US/CT (if obtained) Negative Negative/positive Positive
Persistent drainage No Usually yes Yes
(after 3 wk)b
Reoperation No No Yes
Death related to POPF No No Possibly yes
Signs of infections No Yes Yes
Sepsis No No Yes
Readmission No Yes/no Yes/no
a
Partial (peripheral) or total parenteral nutrition, antibiotics, enteral nutrition, somatostatin analog,
and/or minimally invasive drainage.
b
With or without a drain in situ.
US, ultrasonography; CT, computed tomography scan; POPF, postoperative pancreatic fistula.
Taken from Bassi C, Dervenis C, Butturini G, et al. Postoperative pancreatic fistula: an international
study group (ISGPF) definition. Surgery 2005;138(1):8–13.
diameter), poor nutritional status (albumin <3 g/dL), and high intraopera-

tive blood loss (>1 L). Pancreatic surgeons have also examined different
surgical techniques to avoid fistula formation. For PD, the technique used
(e.g., duct to mucosa vs. invagination; routine placement of a pancreatic
duct stent) or location (pancreaticojejunostomy (PJ) vs. gastrojejunostomy)
of the pancreatic anastomosis has been examined. The results of numer-
ous randomized trials and meta-analyses are mixed, with one favoring one
technique or location over the other, or no difference observed. We rou-
tinely perform a partial invagination technique without the use of stents
and have observed an overall fistula rate of less than 15% and a rate of about
7% after PD for pancreatic cancer.
For distal pancreatectomy, the incidence of pancreatic fistula ranges
from 15% to 25%. Some studies suggest that the fistula rate is decreased with
the use of stapled closure. More recently, there is some evidence that rein-
forced stapled closure, using a Seamguard, may further decrease the rate
of fistula development. We routinely use stapled closure and Seamguard
where possible but have found that even the 4.8-mm staples cannot com-
pletely seal a thick pancreas. In that case, the pancreas is transected with
cautery, and a sutured closure of the cut surface is performed with a run-
ning and locking 3-0 Prolene suture and an additional suture (4-0 or 5-0
Prolene) to close the pancreatic duct. The incidence of pancreatic fistula is
highest in patients undergoing middle (segmental) pancreatectomy rang-
ing from 20% to 60%. This high incidence is due to the presence of both a
closed pancreas surface toward the head of the pancreas and a pancreati-
cojejunal anastomosis of the distal segment.
Treatment
Pancreatic fistulas are associated with increased patient morbidity, hospi-
tal stay and costs, and even occasional death due to sepsis or hemorrhage.
Therefore, prompt diagnosis and treatment are essential. Four goals of
treatment will be reviewed. These include (a) good drainage, (b) treatment
of infection, (c) maintenance of good nutritional status, and (d) correction
of electrolyte imbalances.
002086377.INDD 100 7/14/2014 3:36:03 PM

Chapter 9 / Complications of Pancreatic Surgery 101
The most important treatment goal of pancreatic fistula

anagement focuses on complete drainage of extraluminal pancreatic
m
exocrine juices. We routinely place a closed-suction Silastic drain (10 flat
or 19 round Jackson-Pratt) anterior to the PJ anastomosis. It is drawn
behind the stomach and the left lobe of the liver, which usually maintains
its proper position on top of the PJ suture line. It is cut long enough so
that its tip lies next to the hepaticojejunostomy anastomosis; thus, it will
drain a leak from either site effectively. If the drain output does not appear
turbid or contain bile, the drain is routinely removed on postoperative day
5 to 7, when the patient is eating a regular diet. A high-volume output of
serous fluid does not influence this decision, but if the character of the
fluid is suspicious, the amylase level is checked. If a pancreatic fistula is
present, the drain is left in place. If the patient’s white blood cell count
is elevated, or if they are febrile, broad-spectrum antibiotics are started
and a CT scan performed to look for any undrained fluid collections,
which should be drained by the interventional radiologist with CT scan
or ultrasound guidance (Fig. 9.1). The fluid is sent for culture and amy-
lase concentration, and the antibiotics are then tailored to the specific
organisms. Once all signs of infection resolve, the patient is discharged
with the surgical drain (and any additional percutaneous drains) in place.
Before discharge, around postoperative day 10, the drains are taken off of
suction and allowed to drain into a plastic bag. The patient is seen within
the next week in the clinic. The drains are removed if their output has
been under 10 mL/d for a period of at least 2 to 3 consecutive days. If the
fistula persists several weeks after discharge, we obtain a fistulagram and
at that time replace the original drain with a red rubber catheter (Fig. 9.2).
Occasionally, the tip of the original drain has migrated inside the lumen
of the bowel. If this is the case, the replacement tube is positioned more
superficially. This will hasten closure of the fistula, and the drain can be
removed. If a low-output fistula (<50 mL/d) persists several weeks after
drain exchange, rather than immediately removing the drain, it is backed
out by 3 to 4 cm each week, so the fistula track can heal behind the drain.
Using this strategy, we have never had to reoperate on a patient for a post-
operative fistula after a PD.
The use of somatostatin for the prevention or treatment of postop-
erative pancreatic fistulas has been examined. Overall, it does not decrease
fistula development, hasten its closure, or alter its morbidity. Somatostatin
does decrease the fistula volume, which can help with management of elec-
trolyte deficiencies, nutritional status, and skin breakdown in high-output
fistulas. Short-acting somatostatin is administered as 100 μg subcutane-
ously three times per day; the long-acting variety, which is effective for 1
month, is preferred.
If a fistula persists for 2 to 3 months in patients who undergo distal
pancreatectomy, despite drain exchange as described above, endoscopic
retrograde cholangiopancreatography (ERCP) and a pancreatic duct stent
are indicated. The ERCP can confirm the site of the leak and identify any
downstream ductal strictures that may contribute to the persistence of
the fistula. Even without a stricture, a stent in the main pancreatic duct
that traverses the duodenal papilla decreases the pressure in the duct,
which likely hastens fistula closure. To avoid stent-related complications
(migration, erosion, infection, stricture formation, etc.), the stent must be
removed within 2 months at most.
If, despite all of these measures, the pancreatic fistula does not close
after at least 6 months, then surgical internal drainage may be indicated.
A fistulojejunostomy is created with a Roux-en-Y jejunal limb sewn to the
fistulous tract as close to the pancreas as possible.
0002086377.INDD 101 7/14/2014 3:36:03 PM

B
FIGURE 9.1 Postoperative pancreatic fistula managed with a percutaneous drain. A. CT
scan (coronal views) taken on postoperative day 6 after a pylorus-preserving Whipple resec-
tion for a duodenal cancer reveals a posterior leak from the pancreaticojejunostomy (arrow
highlights intra-abdominal fluid collection). B. Repeat CT scan obtained 5 days after a per-
cutaneous drainage catheter was placed reveals complete resolution of the collection and
the tip of the radiology drain (arrow).
002086377.INDD 102 7/14/2014 3:36:05 PM

FIGURE 9.2 “Tubogram” reveals presence of pancreatic fistula. A tubogram obtained

through a 16-Fr red rubber catheter 3 weeks after a pylorus-preserving pancreatoduode-
nectomy reveals contrast filling the jejunum at the level of the pancreaticojejunostomy
(arrow), confirming the presence of a pancreatic fistula.
For nutritional support, patients with pancreatic fistulas should not

be kept NPO but allowed to eat a regular diet. Studies have shown that
enteric intake is associated with a higher rate of fistula closure. Serum albu-
min levels should be kept at greater than 3 to 3.5 g/dL. Occasionally, supple-
mental calories with TPN are required as an adjunct to oral intake to reach
that nutritional goal. This is the case particularly in patients with nausea
and ileus, which can accompany fistulas and may diminish patients’ appe-
tites. Loss of pancreatic exocrine secretions may also be associated with
electrolyte imbalances, as the fluid is rich in bicarbonate and other essen-
tial electrolytes. Serum electrolyte levels should be routinely monitored in
patients in the hospital and after discharge.
DELAYED GASTRIC EMPTYING

DGE is a common complication after PD. It is defined as high nasogastric
(NG) tube output in the postoperative setting requiring that the tube be left
in place for 10 or more postoperative days, emesis after NG tube removal
necessitating replacement of the tube, or failure to progress to a regular
diet without the use of prokinetics. DGE is not usually a serious event, but
it does prolong hospital stay, temporarily worsens quality of life, and can
occasionally lead to additional invasive procedures.
It is difficult to identify the incidence of DGE after PD, as the defini-
tion in the literature is quite variable. Randomized trials cite a rate from
0% to 57%, with an overall mean incidence of 15%. The International Study
Group on Pancreatic Surgery (ISGPS) published a complex consensus defi-
nition in 2007, to standardize its meaning and enable the comparison of
new approaches aiming to reduce the rate of DGE.
The etiology of DGE is uncertain. Potential causes include decreased
plasma motilin concentrations due to duodenal resection, extended lymph
0002086377.INDD 103 7/14/2014 3:36:06 PM

node dissection along the hepatic artery with disruption of vagal and sym-
pathetic innervation to the antropyloric region, relative devascularization
or denervation of the pylorus after pylorus-preserving PD (PPPD), anasto-
motic disruption at the pancreaticojejunostomy, and transient pancreatitis.
The issue of whether DGE is more common after PPPD than standard PD
has been debated. The results of studies are mixed, with rates of DGE higher
for PPPD in some studies and in PD in others. It is now generally accepted that
both operations are equivalent. The routine administration of postoperative
prokinetics to reduce the rate of DGE (e.g., erythromycin or metoclopramide)
has also not shown benefit. There is evidence that an antecolic duodenojeju-
nostomy (vs. retrocolic) after a PPPD lowers the rate of DGE.
Treatment
Our management of patients with DGE aims to minimize patient burden and
has led to universal cure of the condition with only a small minority of patients
requiring an additional intervention after surgery. We routinely place an NG
tube at the time of surgery, after the patient is under general anesthesia, and
leave it to intermittent suction overnight. It is removed on the first morning
after surgery unless the output is extremely high (>1 L), which occurs rarely.
Patients with DGE usually do not experience symptoms until postoperative
day 5 or later when they develop nausea, abdominal distension, or vomiting.
An NG tube is reinserted in patients with moderate or severe abdominal dis-
tension, vomiting, or an enlarged stomach evident on CT scan. It is placed to
intermittent suction and the amount of output recorded. Patients with DGE
usually have a high daily volume (>1.5 L) of gastric or bilious output. A CT scan
of the abdomen and pelvis with oral and intravenous contrast is obtained to
determine the degree of gastric distension and to identify the presence of an
early postoperative bowel obstruction or undrained intra-abdominal fluid
collections. DGE may be associated with a pancreatic fistula, peripancreatic
fluid collections, or an intra-abdominal abscess. The decision to obtain a CT
scan is not dependent on the presence of an elevated white blood cell count
or fever but rather on the patient’s symptoms as previously described. If fluid
collections are present on the CT scan, and an infection is suspected, ultra-
sound- or CT-guided percutaneous drainage is performed. DGE associated
with undrained and infected fluid collections usually resolves once the fluid
has been drained. If the diagnosis of DGE is still unclear after the CT scan,
an upper gastrointestinal (UGI) contrast series is obtained. This confirms
delayed emptying of contrast from the stomach.
Once the diagnosis of DGE is suspected, patients are kept NPO. The
NG tube decompresses the stomach, which probably hastens its functional
recovery. It is left in place at least for 2 to 3 days until the output is minimal
and the patient can handle his or her own gastric and salivary secretions.
Occasionally, a trial of the NG tube to gravity may be useful.
A promotility agent, usually intravenous metoclopramide, is started
immediately after DGE is diagnosed and a bowel obstruction has been
excluded. Because metoclopramide is associated with numerous side
effects, particularly tardive dyskinesia in elderly patients, all patients
should be closely observed during its administration. If it is needed, eryth-
romycin is an alternate promotility agent that binds motilin receptors and
may be effective as well. Patients can be transitioned from intravenous
metoclopramide or erythromycin to an oral form once their NG tube has
been removed.
A peripherally inserted central catheter (PICC) line is placed early after
diagnosis, particularly in those patients in whom DGE is predicted to last
more than a few days. In these patients, TPN is begun and rapidly compressed
to infuse over a 12-hour period. Home health nursing is arranged to assist in
002086377.INDD 104 7/14/2014 3:36:06 PM

home TPN care. Patients are discharged on TPN and a promotility agent and
allowed to slowly advance their diet at home. They are seen in the outpatient
clinic on a weekly basis until the DGE resolves. Using this strategy, we have
found that almost all cases completely resolve within several weeks.
In rare instances, patients have a more prolonged course of abdominal
distention, nausea, or even intermittent vomiting that can require numerous
hospital readmissions. At each admission, NG tube decompression is repeated.
Occasionally, a percutaneous endoscopic gastrostomy tube has been placed to
minimize the symptoms and allow for gastric decompression at home. In our
experience, reoperation to deal with DGE has not been required.
POSTOPERATIVE HEMORRHAGE
The International Study Group on Pancreatic Surgery (ISGPS) released a
consensus definition on postpancreatectomy hemorrhage in 2007, which
classifies it according to severity and the time of onset after surgery. The
different severity classifications include types A, B, and C with the latter two
involving major hemorrhage with a hemoglobin drop of greater than 3 g/
dL. “Early postoperative hemorrhage” occurs within the first 24 hours of
surgery; “late hemorrhage” occurs after 24 hours. Studies using these defini-
tions cite an incidence of postoperative hemorrhage at 3%.
Prior to the improvement of endoscopic and interventional radio-
graphic angiography techniques, post–pancreatic surgery hemorrhage
(PPH) was associated with an overall mortality rate of 50%. Mortality was
even higher in those patients presenting 72 or more hours after surgery.
The greatest risk of late PPH occurs in patients who develop postopera-
tive pancreatic fistulas from a nonhealing pancreaticojejunostomy anas-
tomosis. Often, these patients also develop vascular abnormalities (e.g.,
pseudoaneurysms) that further worsen the outcome. The observation that
mortality rates have decreased with the development of improved nonop-
erative techniques has led to the overall goal of avoiding repeat surgery in
patients with late hemorrhage. This treatment strategy is in contrast to that
for patients with early postoperative hemorrhage, who should generally
undergo prompt operative reexploration.
Treatment
Postoperative hemorrhage should be suspected in patients whose hemo-
globin decreases by more than 3 g/dL over the first postoperative day or
those who develop systemic evidence of hypovolemia and bleeding. Once
suspected, resuscitative treatment should be initiated rapidly and urgent
steps taken to determine the source of the bleeding, which is either extra- or
intraluminal (i.e., within or outside of the intestinal tract). Almost all bleed-
ing within the first 24 to 48 hours is from an extraluminal source and is often
due to a technical error during the operation. Nevertheless, the coagulation
profile is checked to ensure that there is no evidence of synthetic liver dys-
function, manifested by a high prothrombin time (PT). If the coagulation
parameters are abnormal, fresh frozen plasma is promptly administered to
correct the deficit. Patients who have had prolonged biliary obstruction and
high preoperative bilirubin may also have elevated PT levels due to vitamin
K malabsorption and deficiency. They should be given intravenous vitamin
K if their PT levels are elevated. If the platelet count is less than 50,000 per
dL, platelets should be transfused. However, the surgeon should have a
low threshold for return to the operating room for exploration, when most
patients will be found to have a correctable source of bleeding. Figure 9.3
illustrates the potential bleeding sources.
0002086377.INDD 105 7/14/2014 3:36:06 PM

B
FIGURE 9.3 Common sites of bleeding after (A) pancreaticoduodenectomy or (B) distal
pancreatectomy. (From Wente MN, Veit JA, Bassi C, et al. Postpancreatectomy hemor-
rhage (PPH): an International Study Group of Pancreatic Surgery (ISGPS) definition. Surgery
2007;142(1):20–5. PMID: 17629996.)
Late PPH (>24 hours after surgery) can occur over a long time span that
extends to weeks or even months postoperatively. Therefore, it is important
for the surgeon to be aware of this possibility and to educate the patients as
to the signs and symptoms, since most patients are at home when it occurs.
Unlike early PPH, which is almost always extraluminal, late PPH often is
intraluminal, although bleeding of either type can occur. Patients with intra-
luminal hemorrhage present like those with UGI bleeding; they experience
002086377.INDD 106 7/14/2014 3:36:09 PM

hematemesis or melena. The most likely sources are the three surgical
anastomoses, and the duodenojejunostomy is the most frequent. An NG
tube is inserted to verify that blood is in the stomach. If it is present, a UGI
endoscopy is performed for both diagnostic and therapeutic purposes. If the
bleeding source is identified, cautery is applied or a clip is placed. Patients
must still be closely monitored for the ensuing 48 hours. If the source is not
identified, or if the suspected site of bleeding is an extraluminal site, the next
step is to employ selective angiography both to identify the bleeding site and
to stop the bleeding via various interventional angiographic techniques.
Selective angiograms of the celiac axis and superior mesenteric arteries are
performed (Fig. 9.4). As illustrated in Figure 9.3, the most common sources
of bleeding are the gastroduodenal artery and transverse pancreatic arcades
adjacent to the pancreaticojejunostomy anastomosis.
In rare instances, late PPH cannot be controlled via endoscopic or
angiographic techniques. If this is the case, the patient should be operated
upon urgently. If the source of bleeding is the pancreaticojejunostomy and
there is an associated anastomotic dehiscence, the anastomosis may also
need to be addressed. This can be done with placement of drainage cath-
eters, reinforcement of the anastomosis with additional sutures, or addi-
tional pancreatic resection. However, it should be stressed that surgery
for late bleeding is required infrequently today, when the patient is being
managed in a “center of excellence” where a skilled multidisciplinary group
is available. Reoperative surgery in this setting is still associated with an
extremely high mortality rate.
BILIARY FISTULA
Bile leaks from the choledochal/hepaticojejunal anastomosis occur in 1% to
2% of patients undergoing PD and are heralded by the appearance of bile in
the drain fluid. If this occurs, the drain should be left in place until the leak
stops. If bile is still present when the patient is ready for discharge, they can
go home with the drain in place. It can be removed in the office when there
is no longer any bile present. If there is no evidence of a bile leak, the biliary
drain is removed the day after the patient begins oral intake. Because the PJ
anastomosis is close to the bile duct anastomosis, pancreatic fistula fluid
can also be tinged with bile. Because a fistula from either site is managed in
a similar way, it may not be necessary to determine its origin with certainty.
If this is required, a contrast injection through the drain tube can resolve the
question. Management of a high-volume (>200 mL/d) bile leak is more com-
plex, often requiring percutaneous transhepatic biliary stent placement.
EXOCRINE AND ENDOCRINE INSUFFICIENCY

The incidence of symptomatic exocrine insufficiency after pancreatic sur-
gery depends on the amount of parenchyma resected, the function of the
pancreas that remains, and the adequacy of mixing of the food with pan-
creatic enzymes and bile. The first two indices can be estimated with some
degree of confidence and the last cannot. Thus, we council our patients that
if at least 20% of a normal pancreas remains, it should be enough to provide
adequate digestive (and endocrine) activity. This is a reliable prediction for
patients who will undergo a distal pancreatectomy, including removal of the
entire body and tail of gland, but it is less so for patients who will undergo
a pancreatoduodenectomy. To some degree, this is because many patients
with pancreatic tumors that have obstructed the main pancreatic duct
may have obstructive chronic pancreatitis in the remnant pancreas, lead-
ing to decreased exocrine function. But it is mostly due to the uncertainty
about how well ingested food will mix with the pancreatic secretions and
0002086377.INDD 107 7/14/2014 3:36:09 PM

B
FIGURE 9.4 Gastroduodenal artery pseudoaneurysm treated angiographically. A. A supe-
rior mesenteric artery angiogram reveals a gastroduodenal artery pseudoaneurysm from a
completely replaced hepatic artery (arrow). B. A 5 × 50 Viabahn stent placed across the
pseudoaneurysm with complete occlusion of the bleeding site maintaining perfusion to the
common hepatic artery distal to the pseudoaneurysm.
002086377.INDD 108 7/14/2014 3:36:11 PM

bile in the small intestine. This is true for both a standard PD as well as the
pylorus-preserving variety; studies show that after either operation, gastric
emptying may be normal or faster or slower than normal. So the amount of
pancreatic enzymes secreted by the remnant pancreas may be adequate,
but if those enzymes do not mix well with the food, they cannot be effective.
It is likely that the majority of patients have some degree of malab-
sorption and steatorrhea (i.e., >7% excretion of ingested fat) after any major
pancreatic resection. We do not treat them unless there are symptoms of
weight loss or inability to gain weight, diarrhea, oily stools, and/or abdomi-
nal bloating. Treatment consists of supplemental enzymes with at least
30,000 IU of lipase taken with each meal; half that amount should be taken
with snacks. Only rarely does fat intake need to be limited. All of our PD
patients also are given proton pump inhibitors permanently to minimize
the likelihood of marginal ulceration. It is important to make certain that
these are given as well to those patients who have had other types of pan-
creatic resection and who require pancreatic enzymes, especially if a non-
enteric preparation is used. This is because gastric acid can denature the
ingested enzymes, which makes them ineffective. If patients continue to be
symptomatic, they are referred to a gastroenterologist with experience in
the treatment of pancreatic exocrine insufficiency.
The incidence of endocrine insufficiency (type I diabetes mellitus, DM)
depends mostly on the amount of pancreatic parenchyma removed. Its inci-
dence generally parallels the occurrence of pancreatic exocrine insufficiency
for each specific pancreatic operation. The maintenance of gastroduodenal
continuity also may be an important factor for the development of DM
because it maintains neurohumoral connections between the stomach, the
duodenum, and the pancreas. Thus, the development of diabetes appears
to be less likely after duodenum-preserving pancreatic resections done for
chronic pancreatitis (Beger, Frey), compared to a pylorus-preserving PD.
Patients with preexisting DM can experience worsening of their dis-
ease with pancreatic resections. Early in the postoperative period after any
type of major surgery including pancreatic resection, blood sugars are gen-
erally elevated. Later, if diabetes is suspected after patients resume a diet,
we request an endocrinology and dietary consult, and subsequent man-
agement of DM is supported by those individuals and, eventually, by the
patient’s internist. Patients should also be made aware of the signs/symp-
toms of hypoglycemia, since with the resection of alpha islet cell mass and
less available glucagon, the DM may be more brittle than usual.
Suggested Readings
Bassi C, Dervenis C, Butturini G, et al. Postoperative pancreatic fistula: an interna-
tional study group (ISGPF) definition. Surgery 2005;138(1):8–13. PMID: 16003309.
Correa-Gallego C, Brennan MF, D’Angelica MI, et al. Contemporary experience with
postpancreatectomy hemorrhage: results of 1,122 patients resected between 2006
and 2011. J Am Coll Surg 2012;215(5):616–621. PMID: 22921325.
Kazanjian KK, Hines OJ, Eibl G, et al. Management of pancreatic fistulas after
pancreaticoduodenectomy: results in 437 consecutive patients. Arch Surg
2005;140(9):849–854; discussion 854–856. PMID: 16172293.
Wente MN, Bassi C, Dervenis C, et al. Delayed gastric emptying (DGE) after pancreatic
surgery: a suggested definition by the International Study Group of Pancreatic
Surgery (ISGPS). Surgery 2007;142(5):761–768. PMID: 17981197.
Wente MN, Veit JA, Bassi C, et al. Postpancreatectomy hemorrhage (PPH): an
International Study Group of Pancreatic Surgery (ISGPS) definition. Surgery 2007;
142(1):20–25. PMID: 17629996.
Yekebas EF, Wolfram L, Cataldegirmen G, et al. Postpancreatectomy hemorrhage:
diagnosis and treatment: an analysis in 1669 consecutive pancreatic resections.
Ann Surg 2007;246(2):269–280. PMID: 17667506.
0002086377.INDD 109 7/14/2014 3:36:11 PM

SECTION II
Liver
002086378.INDD 110 7/14/2014 6:51:33 PM

10 Liver Anatomy, Physiology,
and Preoperative Evaluation
Ching-Wei D. Tzeng and Jean-Nicolas
Vauthey
INTRODUCTION
Surgical resection remains the first-line treatment of appropriately selected
patients with primary and metastatic liver cancer. Until the last two
decades, hepatectomies were considered operations of prohibitive risk with
frequent hemorrhagic events and high rates of complications and death.
Even as surgeons have pushed the limits with resectability criteria, the
overall mortality rate has fallen significantly from 10–20% to around 2.5%
nationally. This mortality decrease is the result of more advanced surgical
techniques, better patient selection and risk stratification, optimization
of the future liver remnant (FLR) (including the use of portal vein embo-
lization [PVE] and limiting preoperative chemotherapy), multidisciplinary
treatment strategies, and better diagnostic and therapeutic capabilities to
rescue patients after complications. This chapter focuses on the preopera-
tive considerations a surgeon must evaluate before performing a hepatic
resection.
ANATOMY
In 1888, Rex described the anatomy of the portal vein and the main plane
dividing the left and right liver along the middle hepatic vein. His detailed
illustrations based on embryology set the stage for modern segmental
anatomy (Fig. 10.1). In 1954, Couinaud divided the liver into four sectors
based on the right, middle, and left fissures, which follow the vertical
planes of the three hepatic veins (Fig. 10.2). They include the right pos-
terior (segments VI + VII), right anterior (segments V + VIII), left medial
(segments III and IV), and left lateral (segment II) sectors. The transverse
scissura follows the plane of the portal vein, which then divides the sectors
into each segment.
The International Hepatopancreatobiliary Association (IHPBA) con-
sensus Brisbane (2000) classification for liver surgery terminology is now
widely adopted, with the liver divided into sections rather than sectors.
The main difference is that segments II and III now belong together as
the left lateral section with segment IV remaining as the medial segment
without segment III (Fig. 10.3). In an effort to standardize terminology
among surgeons and surgical literature, the terms right/left lobectomy,
bilobar, and trisegmentectomy are now mostly replaced with the terms
right/left hepatectomy, bilateral, trisectionectomy, sectionectomy, and
segmentectomy. Detailed knowledge of the segmental and vascular anat-
omy is critical to preoperatively planning the amount of liver parenchyma
that will need to be resected and how much FLR will remain after resec-
tion. Precise terminology is critical for communication among surgeons,
radiologists, and the multidisciplinary team.
111
0002086378.INDD 111 7/14/2014 6:51:33 PM

112 Section II / Liver
FIGURE 10.1 Intrahepatic portal vein anatomy as depicted by Rex in 1888.
PHYSIOLOGY
For safe resection, the liver parenchyma must have sufficient size, adequate
synthetic function, unimpaired biliary drainage and perfusion, and suffi-
cient regenerative capacity. A global assessment of parenchymal quality is
important to determine the functional ability of the liver to tolerate resec-
tion and to guide the surgeon in choosing a safe maximum volume that can
be resected. This can involve the patient’s history, physical exam, biochemi-
cal tests, and radiologic imaging.
For patients with cirrhosis, the Child-Pugh score classifies the sever-
ity of chronic liver disease. This score incorporates albumin, ascites, bili-
rubin encephalopathy, and international normalized ratio (INR) and, by
using a scoring system, stratifies three classes of mortality risk (Table 10.1).
Usually, only Child-Pugh A patients are considered operable because the
002086378.INDD 112 7/14/2014 6:51:35 PM

Chapter 10 / Liver Anatomy, Physiology, and Preoperative Evaluation 113
FIGURE 10.2 Liver segments numbered clockwise from I to VIII and sectors (dotted lines)
along the planes of the three hepatic veins as described by Couinaud. (Adapted from
Couinaud C. Lobes et segments hepatiques: notes sur l'architecture anatomiques et chirur-
gicale du foie. Presse Med 1954;6(2):709–712.)
pathophysiologic consequences of cirrhosis are minimal, and thus they are

described as “compensated.” Some surgeons will operate on Child-Pugh B
patients but with a known increased perioperative risk. The operative mor-
tality of Child-Pugh C patients can approach 50% even for “minor” general
surgery operations such as umbilical hernia repair and cholecystectomy.
Child-Pugh C patients cannot tolerate any liver resection. The best treat-
ment for hepatocellular carcinoma (HCC) in patients with Child-Pugh B
and C cirrhosis is liver transplantation if they qualify.
An alternative system for grading cirrhosis is the Model for End-Stage
Liver Disease (MELD), which consists of bilirubin, creatinine, INR, and liver
disease etiology, in a formula originally created to predict 3-month mor-
tality after a transjugular intrahepatic portosystemic shunt (TIPS) proce-
dure and validated for predicting mortality in hospitalized and ambulatory
patients awaiting liver transplant. As an alternative to Child-Pugh score,
MELD score can be used to predict perioperative and long-term mortal-
ity after resection of HCC. While patients with high MELD scores (≥ 9) are
inoperable for surgical procedures other than transplant, those with MELD
scores less than 9 can be considered for hepatectomy.
While these rules of thumb are easy to follow if a diagnosis of cirrhosis
already exists, the real clinical problem occurs when chronic liver disease
is unrecognized until surgical exploration. Thus, patients should be care-
fully screened preoperatively for cirrhosis and portal hypertension based on
history (jaundice, hepatitis, upper gastrointestinal [GI] bleeding, encepha-
lopathy), physical examination (skin varices or superficial venous collater-
als, ascites, mental examination), serum laboratory tests (bilirubinemia,
thrombocytopenia, anemia from bleeding), and imaging (ascites, spleno-
megaly, scalloping of the liver, left lateral section and caudate hypertrophy,
intra-abdominal varices, and venous collaterals). In general, patients with a
bilirubin greater than 2 mg/dL are not considered for resection. In patients
who present with obstructive jaundice, resection is only considered after
resolution of the jaundice using preoperative endoscopic or percutaneous
biliary drainage.
Because there is a sliding scale of FLR volume minimums and liver
quality, liver biopsy can be used to diagnose preexisting liver damage and
0002086378.INDD 113 7/14/2014 6:51:35 PM

FIGURE 10.3 Resection planes using the International Hepatopancreatobiliary Association

(IHPBA) Consensus Brisbane (2000) classification for liver surgery. (Adapted from Abdalla EK,
Denys A, et al. Total and segmental liver volume variations: implications for liver surgery.
Surgery 2004;135(4):404–410.)
TABLE
10.1 Child-Pugh score for cirrhosis
1 point 2 points 3 points Class

Albumin (g/dL) > 3.5 2.8–3.5 < 2.8 A: 5–6 points
Ascites None Yes, controlled Yes, refractory B: 7–9 points
C: 10–15 points
Bilirubin (mg/dL) < 2 2–3 > 3
Encephalopathy None Grade I–II Grade III–IV
(controlled) (refractory)
INR < 1.7 1.71–2.20 > 2.20
INR, international normalized ratio.
From Child CG, The liver and portal hypertension. Philadelphia, PA: Saunders, 1964.
002086378.INDD 114 7/14/2014 6:51:36 PM

FIGURE 10.4 Diagram of the future liver remnant required for hepatectomy depending on
the parenchymal quality and any preexisting liver injury. (BMI, body mass index).
evaluate whether this damage is acute, subacute, or chronic in nature

(Fig. 10.4). Preoperatively, this can be easily done percutaneously. Or, if
diagnostic laparoscopy is being performed for staging or for another minor
procedure, a wedge biopsy can be done simultaneously. However, as sim-
ple as a biopsy may seem, two major problems must be recognized. First,
the distribution of parenchymal diseases, such as fibrosis, steatosis, and
chemotherapy-associated liver injury (CALI) such as sinusoidal obstruc-
tion syndrome (SOS) and steatohepatitis (SH), is heterogeneous, and thus
a small liver biopsy can be associated with false-negative results. Secondly,
these histopathologic diagnoses do not accurately predict postresection
liver function, FLR regenerative capacity, and protection from postopera-
tive hepatic insufficiency (PHI). Thus, while a positive liver biopsy may be
helpful, caution should still be exercised if the biopsy is “normal” but the
liver looks clinically damaged. Some parenchymal changes such as fatty
liver can be grossly evaluated with cross-sectional imaging by comparing
the tissue perfusion and attenuation to that of the spleen and abdominal
fat. Many cirrhotic changes from portal hypertension are also readily appar-
ent on imaging as aforementioned. Current work is being done to more
precisely correlate computed tomography (CT) and magnetic resonance
imaging (MRI) findings with pathologic parenchymal changes.
Indocyanine green retention rate at 15 minutes (ICG R15) is used
as a global measure of liver function. In this test, faster clearance (lower
retention percentage at 15 minutes) implies better liver function and
drainage. Used extensively in East Asia but less so in North and South
America and Europe, a decision tree known as the Makuuchi criteria is
used to evaluate livers with possible chronic disease before minor hepa-
tectomy (usually for HCC). The qualitative cutoff values of ICG R15 may be
less helpful in stratifying livers for major hepatectomy when compared to
regeneration criteria such as those provided by PVE (see below). ICG R15
cutoffs for wedge resection and segmentectomy are 30% to 39% and 20%
to 29%, respectively. Left hepatectomy or sectionectomy requires 10% to
19% retention or less. Finally, right or extended hepatectomy is consid-
ered only if ICG R15 is less than 10%. An important initial consideration
in applying these criteria is that patients with ascites and/or elevated bili-
rubin (> 2 mg/dL) do not qualify for any type of resection irrespective of
the ICG cutoffs.
0002086378.INDD 115 7/14/2014 6:51:37 PM

PREOPERATIVE EVALUATION
Risk Factors for Complications
Complications specific to liver surgery include bleeding (both intraopera-
tive and early postoperative), bile leak, and postoperative hepatic insuffi-
ciency (PHI or liver failure) with PHI-related mortality. The current authors
previously defined PHI as a peak total bilirubin greater than 7 mg/dL, which
has a sensitivity and specificity of greater than 93% and an odds ratio of
10.8 for predicting 90-day mortality. Acute PHI often leads to an irreparable
slide toward liver failure–related mortality that continues well past 30 days.
In regard to measuring surgical outcomes, one-third of posthepatectomy
deaths occur between 30 and 90 days, so typical short-term metrics fail to
fully capture late liver-related mortalities.
Predictors of major complications can be divided into three catego-
ries—medical comorbidities, laboratory abnormalities, and perioperative
risk factors (including extent of hepatectomy). These parameters include
low albumin, smoking, American Society of Anesthesiologists (ASA) class,
elevated alkaline phosphatase, elevated partial thromboplastin time (PTT),
extent of hepatectomy, prolonged operative time, and intraoperative or
postoperative transfusions. The last three factors are associated with each
other, and thus, surgeons should be cautious with patient selection when
pairing major simultaneous operations with major hepatectomies. While
not every risk factor is reversible, many are potentially modifiable in the
preoperative period through optimization of medical issues and planning
operations of lesser magnitude when oncologically practical.
Initial Assessment
The initial preoperative assessment of a patient requiring a liver resection
takes into account three major factors—patient operability (performance
status and comorbidities), liver quality (global function), and tumor extent
(balance of oncologic resectability and FLR). An irreversible or unmodifi-
able issue with the patient or the liver precludes liver resection. The risks
of liver resection and PHI are relatively unique, because the liver is the only
abdominal visceral organ without which a patient cannot survive. Thus,
unless a transplant is planned, enough functional parenchyma must be left
behind, or the patient will inevitably die from liver failure.
Patient Operability
The term “operability” describes patient, not tumor, factors. Patients can
be operable, inoperable, or borderline operable if they have potentially
reversible comorbidities. Operability is dependent on a patient’s func-
tional capacity (performance status) and underlying medical comorbidi-
ties. If both are good, then the patient is operable. If either is irreversibly
poor, then the patient is inoperable. If functional or medical issues are
potentially reversible, then the patient is considered borderline oper-
able. The fact that grading performance status is part of the National
Comprehensive Cancer Network (NCCN) preoperative assessment for
hepatocellular carcinoma (HCC) highlights the importance of assessing
patient operability. With aggressive prehabilitation, nutritional coun-
seling, and/or medical optimization, a borderline operable patient may
become an operable surgical candidate and has the potential to enjoy the
same survival benefit from resection as a patient who was initially opera-
ble. If a patient is undergoing preoperative chemotherapy, this time period
offers an opportunity to address the aforementioned issues. Even if pre-
operative therapy is not considered or not indicated, taking appropriate
time to optimize patient physiology before surgery will not only decrease
002086378.INDD 116 7/14/2014 6:51:37 PM

surgical complications but will also improve the clinical rescue rate of
patients after major complications and decrease the postcomplication
mortality rate.
Assessment and Optimization of Future Liver Remnant

The FLR is the predicted volume of remaining liver after resection. Sufficient
size, synthetic function, biliary drainage, perfusion, and regenerative capac-
ity are essential to avoid both early and late PHI. Evaluating the FLR’s
volume is currently the most reliable approach to predict outcomes for
patients who are candidates for major resection. Several methods for such
evaluation have been described. At the University of Texas MD Anderson
Cancer Center (MDACC), the estimated total liver volume (TLV) is calcu-
lated using a formula that relies on the linear correlation between the TLV
and body surface area (BSA): TLV (in cm3) = − 794.41 + 1,267.28 × BSA (m2).
The “standardized” FLR is then calculated as the ratio of the predicted FLR
volume to the standardized TLV.
In a series of 301 patients without hepatic injury or chronic liver disease
undergoing extended right hepatectomy, a standardized FLR of less than
20% was a risk factor for PHI and 90-day postoperative mortality. Generally,
the required minimum standardized FLR is greater than 20% for normal
livers, greater than 30% for livers with limited parenchymal injury (e.g.,
extended duration chemotherapy), and greater than 40% for permanently
damaged livers (e.g., cirrhosis, Fig. 10.4). Because chemotherapy is used in
the majority of cases of colorectal liver metastases (CLM) and often sev-
eral months of chemotherapy have already been given before the patient
sees a liver surgeon, it is imperative that a liver surgeon understands the
side effects of oxaliplatin and irinotecan, which, along with 5-fluorouracil,
have been the backbone of chemotherapy regimens for CLM since the early
2000s (Table 10.2).
With increased CT volumetry availability, “eyeballing” a CT scan
to estimate FLR is highly discouraged. Even with no liver damage, up to
75% of patients at initial presentation are anatomically too small for an
extended right hepatectomy based on the volume of segments II + III.
Preoperative PVE allows a functional stress test to observe the regenera-
tive capacity of the FLR before, rather than after, a hepatectomy. During
a brief 5 weeks between PVE and hepatectomy, surgeons can evaluate the
degree of hypertrophy (DH; needs to be > 5%) and the kinetic growth rate
(KGR = DH divided by time in weeks). Patients with a KGR less than 2% may
be spared hepatectomies of excessive extent that would leave inadequate
FLR. Not all PVEs result in adequate hypertrophy, and one putative criti-
cism of PVE is inadequate hypertrophy preoperatively. However, with PVE
TABLE
10.2 Chemotherapy-Associated Liver Injury Characteristics
Agent Liver Changes Clinical Sequelae

Oxaliplatin • Sinusoidal dilation • Increased perioperative
• Sinusoidal obstructive transfusions with > 6 mo of
syndrome (SOS) preoperative chemotherapy
• Microvascular damage
Irinotecan • Steatosis • Increased complications
• Steatohepatitis • Increased overall 90-d mortality
(15%) and liver failure–related
mortality (6%)
0002086378.INDD 117 7/14/2014 6:51:37 PM

of the right portal vein and segment IV with the combined use of spherical
microspheres and coils, an FLR increase of 69% has been documented in
the hands of experienced interventional radiologists.
Tumor Extent and Resectability

Besides patient operability, whether a liver tumor can be resected is depen-
dent upon two other major factors—technical resectability and oncologic
feasibility. Technical resectability is the ability to surgically remove all liver
tumors with R0 (negative microscopic) margins while leaving adequate
FLR. Adequate FLR must have regenerative capacity and consist of at least
two functional contiguous liver segments with biliary drainage and vascu-
lar inflow/outflow. Again, tumor resectability should be differentiated from
patient operability, which is the patient’s physiologic and medical ability to
undergo and recover from major abdominal surgery.
Oncologically, resectability is determined by the tumor biology of the
cancer type and the individual patient. For example, while extrahepatic
disease is a contraindication for resection in primary liver cancer, patients
with limited extrahepatic disease in controllable sites (e.g., small lung
metastases or regional lymph nodes) can still benefit from hepatectomy
for CLM. However, because these patients are biologically at higher risk of
recurrence, they need perioperative chemotherapy. Even patients with lim-
ited progression of preexisting CLM while on preoperative chemotherapy,
but whose lesions remain anatomically resectable, should undergo resec-
tion. However, patients who develop new lesions or interval extrahepatic
disease while on preoperative chemotherapy should not undergo surgery
until the systemic disease is controlled.
Preoperative Planning with Radiographic Imaging

Four major imaging options are used for liver tumors—ultrasound (US),
CT, MRI, and positron emission tomography/CT (PET/CT). US is inexpen-
sive and reliable, but has been replaced by cross-sectional imaging due to
limited image reproduction, operator variability, and decreased ability to
detect small lesions. However, intraoperative US (IOUS) remains a critically
important surgical tool, because it can find additional tumors not seen on
preoperative cross-sectional imaging in up to 10% of patients, even with use
of improved high-resolution CT and MRI. The greatest value of IOUS is in
guiding the resection in relation to hepatic vascular anatomy and directing
parenchymal transection to balance optimal tumor margins with maximal
FLR preservation.
The current standard for preoperative planning is high-resolution,
contrast-enhanced multidetector CT of the chest, abdomen, and pelvis.
For enhanced liver characterization, MRI abdomen and pelvis with chest
CT is an alternative. CT is generally more useful than MRI for comprehen-
sive imaging for staging, including chest and other abdominal structures.
In general, CLM are hypovascular and more prominent in the portovenous
phase as hypodense lesions. On the other hand, HCC are hypervascular on
the arterial phase with washout on delayed images. Although suboptimal
for general abdominal and chest surveillance, MRI is useful for diagno-
sis and staging of liver tumor burden. Using multiple and newer contrast
agents and dynamic phases, high-resolution (3-tesla) MRI offers increased
lesion characterization over CT, especially for subcentimeter indeterminate
lesions. If preoperative chemotherapy is administered, whichever modality
was chosen pretreatment should be repeated after preoperative treatment
and before liver resection for accurate FLR assessment.
There has been a recent dramatic increase in the use of PET/CT to
identify areas of increased metabolic activity, which are often presumed
002086378.INDD 118 7/14/2014 6:51:37 PM

to be metastases in patients with cancer. Unfortunately, PET can also be

nonspecific, and the matched noncontrast CT adds little information for
equivocal areas, prompting additional workup. At MDACC, PET/CT is
reserved for the detection of occult extrahepatic disease in patients with a
high index of suspicion for metastatic disease and equivocal cross-sectional
imaging (e.g., elevated carcinoembryonic antigen [CEA] with normal CT)
when detection would change the planned treatment strategy.
Percutaneous needle biopsies of suspected liver tumors are unnec-
essary when imaging identifies new lesions with characteristic imaging
features for metastases or primary liver cancer. Needle biopsy may be
appropriate when a benign lesion is suspected and cannot be delineated
noninvasively with MRI and again importantly, only if the treatment plan
would change based on the result.
Radiographic evaluation of treatment response has prognostic value
and frequently determines resectability. Although traditionally measured
by Response Evaluation Criteria in Solid Tumors (RECIST) and modified
RECIST criteria, change in tumor diameter may not provide the best prog-
nostic information. Besides using disease-free interval and tumor factors
from published clinical risk scores, biologic selection can be accomplished
by stratifying patients by the morphologic response on CT after preopera-
tive treatment for CLM. The morphologic response did not correlate with
size-based RECIST measurements, and it was better than RECIST at pre-
dicting major pathologic response. When tumors change from heteroge-
neous consistency and irregular borders to homogenous density, clearly
demarcated borders, and no enhancement, this pattern strongly corre-
lates with pathologic response, margin control, and survival (Fig. 10.5).
Suboptimal morphologic response is an independent predictor of worse
overall survival and thus may offer the surgeon and the patient important
prognostic information when counseling about the potential risks versus
benefits of hepatectomy. Figure 10.5 details the optimal 1 morphologic
response seen in a patient who underwent preoperative chemotherapy for
a very large liver metastasis, which encased the right and middle hepatic
veins. With such good morphologic response, undergoing the risk of an
extended right hepatectomy made oncologic sense in that he was the type
of patient who might benefit the most from resection. After PVE, he under-
went a margin-negative extended right hepatectomy and is alive more than
5 years later.
PREOPERATIVE THERAPY AND DOWNSIZING

When patients present with anatomically unresectable CLM, the clinician
should consider the ability to downsize their lesions to resectability crite-
ria. Effective chemotherapy may achieve this goal in 10% to 20% of initially
unresectable patients, and patients who make it to resection share survival
rates far better than those associated with palliative chemotherapy with
outcomes approaching that of patients with initially resectable CLM. To
convert patients from unresectable to resectable, maximal tumor response
is required (unlike the goal for preoperative chemotherapy for resectable
lesions), frequently involving the addition of targeted agents. With these
agents, the conversion rate from unresectability to resectability can be as
high as 13% to 23%.
After downsizing with chemotherapy and targeted therapy, the sur-
gical goal should be an R0 resection of all known sites of CLM, including
known original sites of any lesions that disappeared with complete radio-
graphic response. A recent “problem” due to improving efficacy of systemic
chemotherapy, patients are sometimes presenting to the liver surgeon with
0002086378.INDD 119 7/14/2014 6:51:37 PM

FIGURE 10.5 Optimal response to preoperative chemotherapy. Despite having a positive

portal lymph node, this patient was alive with no evidence of disease at 5 years. He has an
optimal morphologic (loss of enhancement and well-defined lesion) response during preop-
erative chemotherapy (from A to B). His postoperative scan revealed excellent hypertrophy
after extended right hepatectomy following preoperative portal vein embolization (C).
002086378.INDD 120 7/14/2014 6:51:38 PM

complete radiographic response in small tumors. However, this does not

mean there is complete pathologic response. Approximately 60% of “disap-
peared metastases” will recur if left unresected. One preoperative consid-
eration when treating multiple metastases of different sizes is the need for
fiducial marker placement by interventional radiology for tiny lesions so
that if they disappear, the surgeon will know the original site to resect.
FIGURE 10.6 Preoperative portal vein embolization (PVE). Figure (A) shows preembolization
flow to right portal vein branches. Figure (B) shows stasis of portal vein flow after emboliza-
tion of the right portal vein and segment IV portal vein branches with microspheres and coils.
(Continued)
0002086378.INDD 121 7/14/2014 6:51:39 PM

FIGURE 10.6 (Continued) Figures (C) and (D) show the resultant hypertrophy from pre- to
postembolization.
Timing of Surgery and Treatment Sequencing Strategy

A key modifiable risk factor for posthepatectomy complications is the duration
of preoperative chemotherapy, most often used for CLM. Extended-duration
(4 months) chemotherapy only increases the risk of CALI and associated post-
operative complications without improving the pathologic response. Thus,
every patient with CLM who might be a candidate for surgery (including
patients with “initially unresectable” CLM who have bilateral disease, small
livers, or resectable extrahepatic disease) should be evaluated at diagnosis by a
multidisciplinary team that includes a liver surgeon. The goal of chemotherapy
for resectable lesions is not and should not be complete disappearance of the
CLM. Further chemotherapy can be given after recovery from hepatectomy.
002086378.INDD 122 7/14/2014 6:51:40 PM

A second method of decreasing morbidity is treatment sequencing

for synchronous CLM with intact primary colorectal cancer. The liver-first,
or reverse, approach allows the cleaner operation to be done before resec-
tion of the primary tumor while removing the patient’s metastatic disease.
Used for patients with nonobstructing primary tumors, this is usually cho-
sen for patients requiring preoperative chemotherapy and larger hepatic
resections for CLM, which are the real life-limiting process (rather than
the primary lesion) for the patient. Concerns about the morbidity of major
colorectal operations, even proctectomies, staged after major hepatecto-
mies should not preclude reverse approach sequencing.
Frequently, patients with extensive bilateral liver disease require two-
stage hepatectomy. The first stage involves clearing the left liver of small-
volume disease. The time in between stages is used for interval percutaneous
PVE to grow the FLR (Fig. 10.6). If there will be a significant break in ther-
apy, one or two rounds of chemotherapy can be given without any negative
consequence on FLR hypertrophy. The second stage is often an extended
right hepatectomy to clear the remaining tumor burden. While this staged
sequencing is more time and resource consuming compared to single-stage
extended hepatectomy with contralateral ablation or wedge resections, it
provides a greater margin of safety for avoidance of PHI-related mortality.
Figure 10.6 demonstrates the right portal venography before and after PVE
with CT images before and after PVE-induced hypertrophy. With periopera-
tive chemotherapy, two-stage hepatectomy ± PVE can result in prolonged
5-year overall survival of 51% (Fig. 10.7).
FIGURE 10.7 Overall 65% survival in patients with advanced bilateral colorectal liver
metastases (median 6 metastases per patient) treated with chemotherapy and two-stage
hepatectomy. This survival was achieved in a subset of patients with advanced metastatic
disease by selecting patient candidate for surgery based a combination of response to
chemotherapy and adequate regeneration from portal vein embolization. (Reproduced from
Brouquet A, Abdalla EK, et al. High survival rate after two-stage resection of advanced
colorectal liver metastases: response-based selection and complete resection define out-
come. J Clin Oncol 2011;29(8):1083–1090.)
0002086378.INDD 123 7/14/2014 6:51:40 PM

CONCLUSIONS
Over the past two decades, hepatectomy has evolved into a safe and effective
therapy for a wide range of benign and malignant diseases. Postoperative
complications may be related to patient factors, anatomic factors associ-
ated with resection extent, or technical factors that result in major intra-
operative bleeding. Some patient-related factors (e.g., age, Child-Pugh
class, and body mass index) cannot be modified preoperatively. Others can
be reversed with prehabilitation and aggressive medical optimization to
increase the rescue rate of expected complications. The FLR volume and
the degree of hypertrophy after preoperative PVE are important predictors
of outcome and can help select appropriate patients for major hepatecto-
mies. Bleeding can be minimized with preoperative anatomic evaluation,
image-guided resection, low central venous pressure fluid management,
and judicious use of modern surgical instruments. Proper preoperative
evaluation and surgical planning before hepatectomy can minimize surgi-
cal complications, PHI, and mortality.
Suggested Readings
Abdalla EK, Denys A, et al. Total and segmental liver volume variations: implications
for liver surgery. Surgery 2004;135(4):404–410.
Adam R, Delvart V, et al. Rescue surgery for unresectable colorectal liver metastases
downstaged by chemotherapy: a model to predict long-term survival. Ann Surg
2004;240(4):644–657.
Aloia T, Sebagh M, et al. Liver histology and surgical outcomes after preoperative che-
motherapy with fluorouracil plus oxaliplatin in colorectal cancer liver metastases.
J Clin Oncol 2006;24(31):4983–4990.
Aloia TA, Fahy BN, et al. Predicting poor outcome following hepatectomy: analysis
of 2313 hepatectomies in the NSQIP database. HPB (Oxford) 2009;11(6):510–515.
Brouquet A, Abdalla EK, et al. High survival rate after two-stage resection of advanced
colorectal liver metastases: response-based selection and complete resection
define outcome. J Clin Oncol 2011;29(8):1083–1090.
Brouquet A, Mortenson MM, et al. Surgical strategies for synchronous colorectal liver
metastases in 156 consecutive patients: classic, combined or reverse strategy?
J Am Coll Surg 2010;210(6):934–941.
Child CG. The liver and portal hypertension. Philadelphia, PA: Saunders, 1964.
Chun YS, Vauthey JN, et al. Association of computed tomography morphologic crite-
ria with pathologic response and survival in patients treated with bevacizumab
for colorectal liver metastases. JAMA 2009;302(21):2338–2344.
Couinaud C. Lobes et segments hepatiques: notes sur l'architecture anatomiques et
chirurgicale du foie. Presse Med 1954;62:709–712.
Giacchetti S, Itzhaki M, et al. Long-term survival of patients with unresectable colorec-
tal cancer liver metastases following infusional chemotherapy with 5-fluorouracil,
leucovorin, oxaliplatin and surgery. Ann Oncol 1999;10(6):663–669.
Kishi Y, Abdalla EK, et al. Three hundred and one consecutive extended right hepa-
tectomies: evaluation of outcome based on systematic liver volumetry. Ann Surg
2009;250(4):540–548.
Kishi Y, Zorzi D, et al. Extended preoperative chemotherapy does not improve patho-
logic response and increases postoperative liver insufficiency after hepatic resec-
tion for colorectal liver metastases. Ann Surg Oncol 2010;17(11):2870–2876.
Kodama Y, Ng CS, et al. Comparison of CT methods for determining the fat content of
the liver. AJR Am J Roentgenol 2007;188(5):1307–1312.
Kopetz S, Chang GJ, et al. Improved survival in metastatic colorectal cancer is associ-
ated with adoption of hepatic resection and improved chemotherapy. J Clin Oncol
2009;27(22):3677–3683.
Madoff DC, Abdalla EK, et al. Transhepatic ipsilateral right portal vein embolization
extended to segment IV: improving hypertrophy and resection outcomes with
spherical particles and coils. J Vasc Interv Radiol 2005;16(2 Pt 1):215–225.
Makuuchi M, Kosuge T, et al. Surgery for small liver cancers. Semin Surg Oncol
1993;9(4):298–304.
002086378.INDD 124 7/14/2014 6:51:40 PM

Mentha G, Majno PE, et al. Neoadjuvant chemotherapy and resection of advanced

synchronous liver metastases before treatment of the colorectal primary. Br J Surg
2006;93(7):872–878.
Mullen JT, Ribero D, et al. Hepatic insufficiency and mortality in 1,059 noncirrhotic
patients undergoing major hepatectomy. J Am Coll Surg 2007;204(5):854–862; dis-
cussion 862–854.
Rex H. Beitrage zur morphologie der saugerleber. Morph Jahrb 1888;14:517–616.
Ribero D, Abdalla EK, et al. Portal vein embolization before major hepatec-
tomy and its effects on regeneration, resectability and outcome. Br J Surg
2007;94(11):1386–1394.
Rubbia-Brandt L, Audard V, et al. Severe hepatic sinusoidal obstruction associated
with oxaliplatin-based chemotherapy in patients with metastatic colorectal can-
cer. Ann Oncol 2004;15(3):460–466.
Shindoh J, Loyer EM, et al. Optimal morphologic response to preoperative chemo-
therapy: an alternate outcome end point before resection of hepatic colorectal
metastases. J Clin Oncol 2012;30(36):4566–4572.
Shindoh J, Truty MJ, et al. Kinetic growth rate after portal vein embolization pre-
dicts posthepatectomy outcomes: toward zero liver-related mortality in patients
with colorectal liver metastases and small future liver remnant. J Am Coll Surg
2013;216(2):201–209.
Teh SH, Christein J, et al. Hepatic resection of hepatocellular carcinoma in patients
with cirrhosis: Model of End-Stage Liver Disease (MELD) score predicts periop-
erative mortality. J Gastrointest Surg 2005;9(9):1207–1215.
Vauthey JN, Chaoui A, et al. Standardized measurement of the future liver remnant
prior to extended liver resection: methodology and clinical associations. Surgery
2000;127(5):512–519.
Vauthey JN, Pawlik TM, et al. Chemotherapy regimen predicts steatohepatitis and an
increase in 90-day mortality after surgery for hepatic colorectal metastases. J Clin
Oncol 2006;24(13):2065–2072.
Vauthey JN, Zimmitti G, et al. From Couinaud to molecular biology: the seven virtues
of hepato-pancreato-biliary surgery. HPB (Oxford) 2012;14(8):493–499.
Zorzi D, Laurent A, et al. Chemotherapy-associated hepatotoxicity and surgery for
colorectal liver metastases. Br J Surg 2007;94(3):274–286.
0002086378.INDD 125 7/14/2014 6:51:40 PM

11 Cirrhosis and Portal
Hypertension
Rachelle N. Damle and Shimul A. Shah
ETIOLOGY
Cirrhosis is a chronic liver disease that results in hepatic failure. Etiologies
of cirrhosis range from alcohol abuse to immune disorders to toxin expo-
sure. Regardless of its various causes, cirrhosis is caused by the destruction
of hepatic parenchyma with replacement by fibrosis and regenerative nod-
ules. Usually, the damage is done by multiple insults of various etiologies
that occur repeatedly over time. In addition to hepatic failure, cirrhosis can
lead to hepatocellular carcinoma (HCC). The complications of cirrhosis are
devastating and potentially life threatening. Unfortunately, once cirrhosis
occurs, the only cure is liver transplantation (LT).
Pathophysiology
Figure 11.1 details the evolution of cirrhosis. Hepatocytes are damaged by
an insult or disease, leading to destruction of hepatic parenchyma. Initially,
the liver’s ability for regeneration can lead to restoration of normal archi-
tecture; however, this capacity is finite. The exact mechanism leading to
impaired regeneration is poorly understood and remains a key question
in current clinical investigation. Impaired regeneration results in disrup-
tion of architecture, disturbance to blood supply, poor nutrient supply to
hepatocytes, and eventual inhibition of hepatocyte proliferation. Attempts
at healing this damage lead to fibrosis and eventual regenerative nodule
formation.
Stellate cells play an important role in the development of cirrhosis.
They are found at the interface between the basolateral membranes of
hepatocytes and sinusoidal endothelial cells, in the space of Disse. Their
normal function is paracrine in nature, regulating hepatocyte and endothe-
lial cells as well as the storage of vitamin A. When hepatocytes are injured,
cytokine release results in hypertrophy and proliferation of stellate cells.
Eventually, the space of Disse becomes thickened with collagen deposits,
and the normal architecture of the fenestrated sinusoidal endothelium is
distorted. This leads to vascular distortion and portal hypertension.
Under normal circumstances, portal vein pressure ranges from 5 to
8 mm Hg. Any pressure greater than 8 mm Hg or a hepatic portal venous
gradient (HPVG) greater 5 mm Hg is defined as portal hypertension.
Practically, this measurement is taken via hepatic vein wedge pressure,
similar to a measurement of pulmonary arterial pressure using a Swan-
Ganz catheter. The normal venous drainage of the gastrointestinal tract is
through the portal vein, into the liver, out of the hepatic veins, into the infe-
rior vena cava (IVC), and then back to the heart. The portal vein is formed
by the splenic vein and superior mesenteric vein (SMV). Usually, the inferior
mesenteric vein (IMV) joins the splenic vein prior to its junction with the
SMV; however, this anatomy varies widely. With obstruction of flow at the
presinusoidal, sinusoidal, or postsinusoidal level, portal venous pressure
126
002086379.INDD 126 7/14/2014 6:55:28 PM

Chapter 11 / Cirrhosis and Portal Hypertension 127
FIGURE 11.1 Evolution of cirrhosis. Fibrosis develops in nonregenerative necrotic areas,

producing scars. The pattern of nodularity and scars reflects the type of response to injury
(e.g., uniform vs. nonuniform necrosis) and the extent of injury. (From Mulholland MW,
Lillemoe KD, Doherty GM, et al. Greenfield’s surgery: scientific principles & practice, 5th ed.
Philadelphia, PA: Lippincott Williams & Wilkins, 2010.)
increases and decompression occurs with flow back to the heart via col-
laterals. These collaterals are generally thin walled and fragile and, with
increased flow, become varicosities. Given their fragility, varicosities rup-
ture easily and can cause life-threatening bleeding. Figure 11.2 depicts the
collaterals that can form with portal hypertension.
Causes
Table 11.1 lists the causes of cirrhosis. The most common cause of cirrhosis
in the world is viral hepatitis, but alcohol abuse is the most common cause
in the United States. Nonalcoholic fatty liver disease (with progression to
nonalcoholic steatohepatitis or NASH) is increasing in prevalence in the
United States and is linked with hyperlipidemia, non–insulin-dependent
diabetes, and obesity. Immune or inflammatory cirrhosis due to cholesta-
sis is caused by primary biliary cirrhosis and primary sclerosing cholan-
gitis. Hemochromatosis, an inborn error in metabolism, leads to cirrhosis
secondary to lipid peroxidation of iron deposits in periportal regions of the
liver. Wilson disease is an inherited deficiency in hepatocyte transport of
0002086379.INDD 127 7/14/2014 6:55:30 PM

FIGURE 11.2 Potential venous collaterals that develop with portal hypertension. The veins
of Sappey drain portal blood through the bare areas of the diaphragm and through paraum-
bilical vein collaterals to the umbilicus. The veins of Retzius form in the retroperitoneum
and shunt portal blood from the bowel and other organs to the vena cava. (From Mulholland
MW, Lillemoe KD, Doherty GM, et al. Greenfield’s surgery: scientific principles & practice,
TABLE
11.1
Viral
Causes of Cirrhosis
Alcoholic
Nonalcoholic steatohepatitis
Cholestasis
Extrahepatic
Stones
Extrinsic mass
Strictures
Primary sclerosing cholangitis
Intrahepatic
Primary biliary cirrhosis
Primary sclerosing cholangitis
Cystic fibrosis
Vanishing bile duct syndrome
Genetic disorders
Hemochromatosis
Wilson disease
Cystic fibrosis
α-Antitrypsin deficiency
Glycogen storage disease
Venous outflow occlusion
Budd-Chiari syndrome
Idiopathic
002086379.INDD 128 7/14/2014 6:55:32 PM

copper into biliary tract, resulting in copper accumulation in the liver, once
again causing periportal inflammation and eventual cirrhosis. Obstruction
of hepatic venous outflow (Budd-Chiari syndrome) causes sinusoidal con-
gestion and hepatocyte necrosis and, eventually, cirrhosis. Hepatic venous
outflow obstruction is associated with “nutmeg” appearance of the liver sec-
ondary to multiple small areas of hemorrhage.
DIAGNOSIS
As with any disease, a history and physical examination yield most of the
information needed to diagnose a patient with cirrhosis. Clues in the his-
tory include acknowledged alcohol abuse, hepatitis, toxin exposure, previ-
ous upper gastrointestinal bleeding, hemorrhoids, infections, and increased
abdominal girth. Classic physical exam findings in a cirrhotic patient are
found in Table 11.2. Fetor hepaticus, bruising, decreased body hair, and pur-
pura are also physical signs associated with cirrhosis.
Diagnostic tests are indicated if the history and physical exam raise sus-
picion for liver disease. Laboratory evaluation includes hepatic panel, com-
plete blood count, chemistry panel, and coagulation profile. As mentioned
in earlier sections of this handbook, functional capacity of the liver is evalu-
ated by coagulation studies, platelet count, and bilirubin. Inability of the liver
to produce coagulation factors, thrombopoietin, and bilirubin will lead to
abnormal values of these labs. Markers of hepatocyte or cholangiocyte dam-
age are AST, ALT, and alkaline phosphatase. The term liver “function” panel
traditionally includes AST, ALT, alkaline phosphatase, bilirubin, and lactate
dehydrogenase. It is important to understand this misnomer of “function”
as these values do not all indicate the liver’s functional capacity. Importantly,
abnormal values are not specific for liver dysfunction and may even be normal
in the setting of hepatic disease.
In addition to laboratory values, imaging studies are commonly used
to further investigate a suspicion of liver dysfunction. Ultrasonography,
computed tomography, and magnetic resonance imaging (MRI) can detect
cirrhosis. Generally, these studies will reveal a nodular, atrophic liver and
splenomegaly. Ultrasound is nearly 90% sensitive and specific for diagnos-
ing cirrhosis. Cirrhotic livers demonstrate multiple nodular irregularities
TABLE
11.2
Physical Findings
Physical Findings in Cirrhosis
Incidence (%)
Palpable liver 96
Jaundice 68
Ascites 66
Spider angiomas 49
Dilated abdominal wall veins 47
Palpable spleen 46
Testicular atrophy 45
Palmar erythema 24
Noninfectious fever 22
Hepatic coma 18
Gynecomastia 15
Dupuytren contractures 5
From Mulholland MW, Lillemoe KD, Doherty GM, et al. Greenfield’s surgery: scientific principles &
practice. 5th ed. Philadelphia, PA: Lippincott Williams & Wilkins, 2010.
0002086379.INDD 129 7/14/2014 6:55:33 PM

TABLE
Estimated Mortality Rates for Surgical Intervention by CTP
11.3 Score
Child-Turcotte-Pugh (CTP) Score Mortality
A 10%–15%
B 30%–40%
C 75%–90%
on the anterior surface of the liver distinct from the abdominal wall.
Fibrosis can be seen with disruption in parenchyma, though this is often
difficult to detect. Computed tomography and MRI are more expensive
modalities and do not yield any more information than ultrasound; how-
ever, they may reveal an atrophic liver, evidence of venous thrombosis, and/
or splenomegaly.
Invasive diagnostic methods (biopsy) definitively establish the diag-
nosis of cirrhosis; however, in the presence of strong clinical, laboratory,
and radiographic evidence, they are not necessary. Direct visualization
of the liver during an abdominal surgical procedure can reveal a nodular,
atrophic liver. Percutaneous, laparoscopic, or transjugular biopsy, as well as
image-guided fine needle aspiration, can be used. Histology will reveal fatty
infiltration, balloon-cell degeneration, Mallory bodies, hepatocyte necrosis,
fibrosis, or features of cirrhosis.
Two methods of classifying liver disease severity are the Child-
Turcotte-Pugh (CTP) score and the Model for End-Stage Liver Disease
(MELD) score. These prognostic models of disease severity are important
in operative planning and liver transplant allocation. The CTP score is the
best predictive score for most patients with cirrhosis and is used to deter-
mine risk of morbidity and mortality for surgical procedures (Table 11.3).
Originally, the MELD score was developed to determine survival follow-
ing transjugular intrahepatic portosystemic shunt (TIPS) procedure. It
was then adopted as a prognostic indication of a patient’s 90-day sur-
vival with optimized medical management. Bilirubin, international nor-
malized ratio, and creatinine are the only values used in calculating the
MELD score, so it is a more objective system than CTP score. Because of
its objectivity, the MELD score is used as the main determinant of liver
transplant allocation in the United States.
CLINICAL MANIFESTATIONS
Renal
Complications of impaired renal function in the setting of cirrhosis are
caused by dysregulation of vascular tone. This results in sodium retention,
water retention, and ultimately hepatorenal syndrome (HRS) and renal fail-
ure. When ascites develops, there is an inability to excrete sodium. Water
retention follows sodium retention and is due to the inability of patients
with ascites to process free water. Excess water leads to dilutional hypona-
tremia, which may cause nausea, vomiting, lethargy, and seizures.
HRS is a complex complication of cirrhosis, characterized by renal
failure in the absence of intrinsic renal disease. Up to 10% of patients with
cirrhosis and ascites develop HRS. Signs of HRS include oliguria, increasing
serum creatinine level, rising cardiac output, proteinuria, and decreased
arterial pressure. The renin–angiotensin–aldosterone system is overactive,
and the renal cortex is markedly vasoconstricted. Prerenal azotemia is dif-
ficult to distinguish from HRS, and laboratory values are similar in the two
002086379.INDD 130 7/14/2014 6:55:33 PM

TABLE
11.4 Diagnostic Criteria for Hepatorenal Syndromea
Major Criteria
1. Low glomerular filtration rate, as indicated by serum creatinine >1.5 mg/dL or
24-hour creatinine clearance 40 mL/min
2. Absence of shock, ongoing bacterial infection, fluid losses, and current
treatment with nephrotoxic drugs
3. No sustained improvement in renal function (decrease in serum creatinine to
≤1.5 mg/dL or increase in creatinine clearance to 340 mL/min) following diuretic
withdrawal and expansion of plasma volume with 1.5 L of a plasma expander
4. Proteinuria <500 mg/d and no ultrasonographic evidence of obstructive
uropathy or parenchymal renal disease
Additional Criteria
1. Urine volume <500 mL/d
2. Urine sodium <10 mEq/L
3. Urine osmolality greater than plasma osmolality
4. Urine red blood cells <50 per high-power field
5. Serum sodium concentration <130 mEq/L
a
All major criteria must be present for the diagnosis of HRS. Additional criteria are not necessary for
the diagnosis but provide supportive evidence.
From Arroyo V, Ginés P, Gerbes A, et al. Definition and diagnostic criteria of refractory ascites
and hepatorenal syndrome in cirrhosis. Hepatology 1996;23:164, Mulholland MW, Lillemoe KD,
Doherty GM, et al. Greenfield’s surgery: scientific principles & practice. 5th ed. Philadelphia, PA:
Lippincott Williams & Wilkins, 2010.
conditions. However, HRS is distinct from acute intrinsic renal failure, with
low urine sodium, a high urine/plasma creatinine ratio, high urine osmo-
lality, and normal urine sediment. In an attempt to distinguish HRS from
other etiologies, diagnostic criteria for HRS were developed (Table 11.4).
Pulmonary
Although some causes of cirrhosis can also cause pulmonary complications,
cirrhosis itself can have effects on the pulmonary system. As ascites devel-
ops, lymphatic transdiaphragmatic communication can result in hepatic
hydrothorax. The resultant effusion can directly compress pulmonary
parenchyma, which impairs gas exchange and may result in hypoxemia.
Hepatopulmonary syndrome (HPS) is diagnosed with hepatic dysfunc-
tion, oxygen tension less than 70 mm Hg or diffusion gradient greater than
20 mm Hg, and pulmonary vascular dilation in structurally normal lungs.
Patients present with shortness of breath when moving from supine to
upright position (platypnea) and dyspnea in the absence of primary pulmo-
nary disease. On physical exam, patients may have clubbing and cyanosis of
nail beds. Portopulmonary hypertension is defined by a mean pulmonary
arterial pressure of greater than 25 mm Hg in the presence of liver disease.
This rare entity has a significant risk of mortality and is a relative contra-
indication to liver transplant. The exact mechanism of HPS is not known;
interestingly, portal hypertension is not required for its development.
The goal of treating and preventing pulmonary complications of cir-
rhosis and portal hypertension centers around fluid management. With the
previously discussed fluid derangements, extracellular fluid increases and
intravascular volume decreases. Increasing ascitic fluid can impair respira-
tion due to increased intra-abdominal pressure, so therapeutic paracente-
sis can help relieve this pressure and improve air exchange. Prevention of
0002086379.INDD 131 7/14/2014 6:55:33 PM

fluid accumulation is done through sodium restriction. The more complex

treatment of HPS and portopulmonary hypertension is beyond the scope
of this chapter.
Hepatic Encephalopathy
Hepatic encephalopathy is manifested by mental status alterations.
Although a debated topic, the pathophysiology of hepatic encephalopathy
is related to ammonia. Interestingly, the serum concentration of ammonia
does not correlate with the severity of encephalopathy. Normally, after bac-
terial digestion of proteins, ammonia is absorbed through the gut and then
undergoes degradation in the liver. With impaired hepatic function, ammo-
nia reaches the peripheral circulation, permeates the blood–brain barrier,
and causes alterations in mental status. Evidence exists both in favor of and
against elevated serum ammonia as the cause of hepatic encephalopathy.
Serum ammonia levels are elevated in up to 90% of patients with hepatic
encephalopathy, and treatments that reduce ammonia levels improve or
resolve this condition. Alternatively, there is poor correlation of ammonia
level with severity of encephalopathy; direct administration of ammonia to
patients with cirrhosis does not cause encephalopathy; and the same treat-
ments that reduce ammonia levels also reduce other toxin levels. Signs and
symptoms of hepatic encephalopathy include early findings of memory
loss, mood disturbances, and sleep pattern alteration. As the condition pro-
gresses, confusion, lethargy, obtundation, and coma can occur. Asterixis,
elevated ammonia levels, and altered mental status are strongly suggestive
of hepatic encephalopathy.
Treating hepatic encephalopathy involves dietary modifications and
medications that reduce ammonia production and/or neutralize its effects.
Initially, any precipitating factors need to be identified including infectious
source, electrolyte abnormalities, and medications. These factors should be
treated if possible. Secondly, intravenous fluids are administered to expand
intravascular volume and dilute the concentration of the toxins. In addi-
tion to marked dietary restriction of proteins, nitrogenous compounds
should be removed from the gut. This is done via medications that inhibit
the metabolism of such compounds in the gut, thereby reducing ammonia
levels. The preferred medication is lactulose, which can be administered
orally, through a nasogastric tube, or per rectum. Lactulose is administered
three to four times daily, with the goal of producing two to three soft bowel
movements per day. Dosing ranges from 45 to 90 g/d. Its mechanism has
been debated, but is possibly through alteration of bacterial metabolism
to increase bacterial uptake of ammonia. Other medications include neo-
mycin and metronidazole, which work to decrease the concentration of
ammonia-forming bacteria in the gut.
Ascites
Ascites is present when free fluid within the peritoneal cavity reaches over
150 mL. In addition to cirrhosis and portal hypertension, the differential
diagnosis for ascites is broad. Congestive heart failure and renal insufficiency
both commonly cause ascites. Other causes of ascites include hypoalbumin-
emia; bile, chylous, or pancreatic urinary ascites; ovarian disease; perito-
neal dissemination of carcinoma; and myxedema. In the setting of cirrhosis,
increased sinusoidal pressure results in leaking of protein-rich fluid directly
from the liver into the peritoneal cavity. Ascites is of prognostic significance
in a patient with portal hypertension. One-year mortality with new-onset
ascites in a cirrhotic patient is estimated at 50%, compared to 10% in cir-
rhotic patients without ascites. This fluid cannot be absorbed secondary to
the increased hydrostatic pressure and accumulates in the peritoneal cavity.
002086379.INDD 132 7/14/2014 6:55:33 PM

Physical exam findings include shifting dullness, bulging flanks, increased

abdominal girth, and fluid waves. With enough ascites accumulation,
increased intra-abdominal pressure can cause respiratory distress. Increased
intra-abdominal pressure in combination with muscle wasting can result in
abdominal wall and inguinal hernias, which are difficult to manage.
Management of renal sodium and water retention, dietary sodium
restriction, and bed rest are the initial steps in the treatment of ascites.
Sodium retention is exacerbated in the upright position secondary to
venous pooling and relative hypovolemia; therefore, bed rest is recom-
mended; up to 15% of patients respond with this therapy alone. In addi-
tion, dietary salt restriction is implemented; however, compliance with
this recommendation is generally poor. If these lifestyle modifications are
ineffective, diuretics are the mainstay of treatment. Furosemide and spi-
ronolactone are generally prescribed in combination to maximize effec-
tiveness. Intermittent large-volume paracentesis can be performed for
patients with large volumes of fluid that are not responding to lifestyle or
medication interventions. Up to 30 L of fluid can be removed with infusion
of concentrated albumin repletion. Ascites can also be treated with shunts,
either peritoneovenous (surgical) or portosystemic (i.e., TIPS), which are
discussed later in this chapter. Surgical portovenous shunts that are associ-
ated with significant complications are rarely used. Table 11.5 lists treat-
ments of ascites, including doses for antibiotic prophylaxis for spontaneous
bacterial peritonitis (SBP) (see below).
TABLE
11.5
Bed rest
Treatment of Ascites
Sodium restriction
1–2 g/d (45–90 mEq/d)
Fluid restriction
1–1.5 L/d
Diuretics
Spironolactone
50 mg PO q8h
Maximum of 100 mg q6h
Furosemide
40–370 mg/d
Antibiotics
Cefotaxime
2 g IV q12h
Ofloxacin
400 mg PO q12h
Prophylaxis
Norfloxacin
400 mg/d while hospitalized
Ciprofloxacin
750 mg PO weekly
Norfloxacin
400 mg/d for 6 mo
Trimethoprim/sulfamethoxazole
One double-strength tablet five times a week
Mulholland MW, Lillemoe KD, Doherty GM, et al. Greenfield’s surgery: scientific principles & practice.
5th ed. Philadelphia, PA: Lippincott Williams & Wilkins, 2010.
0002086379.INDD 133 7/14/2014 6:55:33 PM

Spontaneous Bacterial Peritonitis

Up to 10% of patients with portal hypertension and ascites will develop
SBP. The etiology of SBP is unclear, but often patients have had preceding
gastrointestinal hemorrhage. Ten to twenty percent of cases are detected
by evaluating routine paracentesis fluid, but most patients present with
clinical symptoms of abdominal pain, fever, progressive encephalopathy,
and impaired renal function. Specific criteria for SBP include a serum–
ascites–albumin gradient of ≥1.1 g/dL, a polymorphonuclear count of ≥250
cells/ mm3, and a single organism (usually G+) on culture. The presence of
more than one organism cultured from peritoneal fluid raises suspicion for
secondary bacterial peritonitis and mandates evaluation for sources such
as a perforated viscus and appendicitis.
Antibiotic prophylaxis against SBP is given to patients with a low
protein count in their ascitic fluid or patients with gastrointestinal hemor-
rhage, as those patients are at high risk for development of SBP. Patients with
hemorrhage are given ofloxacin alone or neomycin, colistin, and nystatin in
combination. These medications can reduce the incidence of SBP from up
to 20% to about 9% with few side effects. Patients with diagnosed SBP are
treated with intravenous cefotaxime or oral ofloxacin.
Esophageal Varices
Esophageal varices are a life-threatening complication of portal hyperten-
sion and can be quite difficult to manage. Esophageal varices are the most
common site for life-threatening variceal bleeding. Typically, they are located
in the distal esophagus, in the submucosal venous plexus. As these vessels
dilate, they become more superficial and they erode into the lamina propria.
Variceal development is dependent upon the pressure in the portal system.
In general, a hepatic vein–portal vein gradient of about 12 mm Hg is present
in patients with varices, although not all patients with this gradient or higher
will develop varices. Incidence varies widely in the literature, ranging from
8% to 90% of patients with cirrhosis. Nevertheless, once varices develop, the
risk of bleeding is 25% to 35%. Predictors of bleeding include a combination
of CTP class, variceal size, and specific markings present on endoscopy.
The mainstay of therapy for esophageal varices is medical, the goal being
decreased splanchnic blood flow and prevention of first bleeding episode.
Beta-blockade is the first-line treatment, with nadolol and propranolol being
the agents of choice. The mechanism of nonspecific beta-blockade includes
decreasing cardiac output as well as increasing splanchnic vasoconstriction;
combined, they result in decreased portal blood flow and decreased hepatic
portal venous gradient. Beta-blockers are effective regardless of the etiology
of portal hypertension, even in the absence of ascites. They can also reduce
mortality from subsequent bleeding episodes in patients with large varices.
Organic nitrates are also effective in primary and secondary prevention of
bleeding. These medications cause splanchnic vasoconstriction and arterial
vasodilation, thereby decreasing collateral resistance. Nitrates also decrease
hepatic resistance, likely a result of stellate cell contractility inhibition. Patients
with a contraindication to beta-blocker treatment receive nitrates as the first-
line therapy. The combination of nitrates with beta-blockers decreases in the
incidence of variceal bleeding to a greater degree than either alone.
MANAGEMENT OF ACUTE VARICEAL BLEEDING

Medical Treatment
Acute variceal bleeding requires prompt attention. Basic management
includes securing a protected airway, establishing intravenous access with
large-bore catheters, close hemodynamic monitoring, thorough laboratory
002086379.INDD 134 7/14/2014 6:55:33 PM

evaluation, intensive care monitoring, and blood product administration

as needed. Medications used in the situation of acute bleeding include
continuous intravenous administration of vasopressin and/or octreotide.
Octreotide infusion is very effective. It is the first-line therapy in acute
variceal bleed, as it results in splanchnic vasoconstriction without any
unintended consequences. Its mechanism is indirect, by reducing the lev-
els of other gut hormones that normally cause vasodilation. Vasopressin
decreases portal venous flow to reduce portal pressure by splanchnic vaso-
constriction. Its vasoconstrictive effects on the coronary circulation can
lead to cardiac complications, so it is often used in combination with a
vasodilator such as nitroglycerin.
Endoscopic Treatment
Endoscopy has been established as a definitive intervention to stop vari-
ceal bleeding through interventions such as sclerotherapy or band liga-
tion. Sclerotherapy involves visualization of the varices through a flexible
endoscope and injection of sclerosing agent in close proximity to the vari-
ces. Sclerotherapy has largely been replaced by band ligation secondary
to its complication rate of 10% to 30%, including esophageal perforation
due to ulceration caused by the sclerosing agent. Success rates for initial
control of esophageal variceal bleed range from 60% to 90%; often more
than one session is required to completely stop bleeding. Band ligation
is performed by insertion of an endoscope over a sheath, suctioning each
individual varix into the lumen of a channel, and placement of a rubber
band around the tissue to ligate it. In a few days, the tissue is sloughed
off, leaving only a small ulcer. Multiple bands can be placed during each
procedure; the exact number is dependent upon the equipment used.
Ligation has a success rate of 80% to 100% in the first session. A signifi-
cant reduction in the incidence of rebleeding, death from bleeding, and
overall mortality has been found with band ligation in comparison to
sclerotherapy.
Balloon Tamponade
Balloon tamponade represents a third-line treatment for patients who
fail medical and endoscopic management for acute variceal bleed. In
the 10% to 25% who fail first-line treatment strategies, balloon tampon-
ade has a high success rate. The Sengstaken-Blakemore and Minnesota
tubes are most commonly used. The concept is similar for both. The
tube has an esophageal balloon and a gastric balloon that sits at the
gastroesophageal junction in the cardia of the stomach. After insertion
via nasogastric route and radiographic evidence of proper position is
demonstrated, the two balloons are inflated, resulting in tamponade
of any gastric or esophageal varices. The esophageal balloon is inflated
between 15 and 40 mm Hg of pressure and the gastric balloon initially
inflated with 30 mL of air and then with a total of 300 to 400 mL. It is
imperative to secure the tube in place, using a facemask or helmet to
prevent inadvertent removal. About 10% to 20% of patients suffer from
significant complications including aspiration, perforation, and necro-
sis; however, 70% to 80% are treated successfully. These severe complica-
tions limit the use of balloon tamponade to 24 hours. Up to half of these
patients rebleed when the balloons are deflated. Balloon tamponade
can be very effective in the initial control of bleeding, but is merely a
temporizing measure until a more definitive procedure such as TIPS or
transplant can be performed. Figure 11.3 shows the algorithm for man-
agement of a patient with variceal bleeding when medical and endo-
scopic treatment fails.
0002086379.INDD 135 7/14/2014 6:55:33 PM

Variceal Bleeding with

Endoscopic/Pharmacologic Failure
Childs A Childs C
Emergency Elective Emergency Elective

Mesocaval Distal splenorenal TIPS Liver Transplant
Portacaval Mesocaval Liver Transplant
TIPS Portacaval
? TIPS Childs B
Emergency Elective
TIPS TIPS
? Liver Transplant Mesocaval
Portacaval
FIGURE 11.3 Suggested treatment options for patients who fail to undergo medical
management for variceal bleeding. (From Mulholland MW, Lillemoe KD, Doherty GM, et al.
Greenfield’s surgery: scientific principles & practice, 5th ed. Philadelphia, PA: Lippincott
Williams & Wilkins, 2010.)
Transjugular Intrahepatic Portosystemic Shunt

Unsuccessful treatment of esophageal variceal bleeding by medication,
endoscopy, or balloon tamponade results in 90% mortality. Therefore,
a more definitive procedure is required that can temporize until LT can be
performed. Traditionally, surgically created portosystemic shunts were the
mainstay of treatment. However, the significant morbidity and mortality
associated with these procedures has led to the development of the less
invasive and lower-risk TIPS procedure. When medical, endoscopic, and
balloon tamponade fail, TIPS is the next line in therapy. This stent effec-
tively shunts blood from the portal vein into the hepatic vein, bypassing
the hepatic parenchyma and avoiding the area of increased resistance.
The goal hepatic portal venous pressure gradient to stop bleeding is less
than 12 mm Hg. However, a gradient of less than 5 mm Hg results in liver
ischemia secondary to decreased hepatic flow. Care needs to be taken
to achieve an adequate pressure to resolve varices, but maintain hepatic
perfusion.
Surgical Shunt Creation

Surgical shunt creation is the most effective method of reducing portal
venous pressure and preventing recurrent variceal bleeding. TIPS is pref-
erential, as the surgical creation of a shunt is associated with significant
morbidity and mortality. Portosystemic shunts are divided into three types
depending on their diversion of flow: nonselective, selective, and partially
diverting.
Nonselective shunts involve anastomosis of the portal vein or SMV
to the IVC or the portal vein to the proximal renal vein. “Hepatofugal
flow” is the reversal of blood flow out of the liver and into the systemic
circulation. It is achieved by the large diameter of these anastomoses and
allows for decompression of the high-pressure portal venous system into
the low-pressure systemic circulation. A portacaval shunt can be done in
an end-to-side or side-to-side fashion. Mesocaval shunts are created via
a prosthetic or vein interposition graft. To create a central splenorenal
002086379.INDD 136 7/14/2014 6:55:35 PM

shunt, a splenectomy is performed and the proximal splenic vein is

anastomosed to the left renal vein. Remember that ascites is caused by
leaking of protein-rich fluid from the liver secondary to high pressure at
the level of the sinusoids. Therefore, these nonselective shunts, by reduc-
tion of portal venous pressure, allow for a decrease in ascites in addition
to resolving varices.
The goal of a selective shunt is to create two distinct drainage path-
ways for the portal circulation. By diverting flow from the esophagogastric
circulation into the systemic circulation, a low-pressure system is created
and maintains a high pressure in the mesenteric system. This is done by
creating a distal splenorenal shunt. Unlike the proximal splenorenal shunt,
the very distal aspect of the splenic vein is anastomosed to the left renal
vein in an end-to-side fashion. This allows decompression of the short gas-
tric veins through the systemic circulation and relieves varices. It is impor-
tant to note that portal venous pressure is not reduced in this procedure.
Therefore, ascites may persist or even worsen after a splenorenal shunt is
created. Another selective shunt involves anastomosis of the left gastric
vein to the IVC and works in the same way as a distal splenorenal shunt.
Partial shunts use small-diameter interposition polytetrafluoroeth-
ylene or Dacron grafts to create mesocaval or portacaval shunts. These
shunts allow for a decrease in the hepatic venous–portal pressure gradient
to less than 12 mm Hg and maintain hepatopetal flow, thereby reducing the
risk of hepatic encephalopathy and accelerated hepatic dysfunction due to
lack of trophic hormones from the gut.
Complications and Outcomes with TIPS and Surgical Shunts

The major complication of TIPS is in-shunt stenosis, via either neointimal
hyperplasia or thrombosis. At 1 year, approximately 50% of shunts will be
occluded, but these results have improved dramatically with newer tech-
nology. Intervention is possible with thrombolysis, balloon dilation of the
stent, or placement of a second stent. About 15% of patients sustain nonre-
versible shunt occlusion. Ultrasonography to assess shunt patency should
be performed every 3 months. Given the relatively short duration of efficacy
of TIPS, it is ideal as a bridge to LT or for those that have severe enough
hepatic decompensation that they are unlikely to live long enough to expe-
rience TIPS failure. TIPS is not only effective in treatment of uncontrolled
variceal bleeding, it also reduces the rate of rebleeding episodes.
Both TIPS and surgically created shunts may be complicated by accel-
erated hepatic dysfunction and hepatic encephalopathy. With diversion of
blood flow from the liver, important trophic hormones present in the portal
circulation bypass the liver. Shunts that involve dissection near the porta
hepatis result in adhesions that make future LT difficult, another reason
that TIPS is preferred over the surgical portosystemic shunts. When it
comes to control of variceal hemorrhage, both nonselective and selective
shunts are successful in greater than 90% of cases.
LIVER TRANSPLANTATION: MANAGEMENT OF PORTAL

HYPERTENSION
The ideal management in selected patients with cirrhosis and at least one
clinical manifestation of portal hypertension is LT. There are currently close
to three times the number of patients listed compared to available organs
in the United States. Due to the shortage of donors and the complexity of
the surgery, donor and recipient selection is key to successful treatment.
Patients typically undergo an exhaustive pretransplant workup to be consid-
ered candidates. Current 1-year survival rate after LT is 85%. Perioperative
0002086379.INDD 137 7/14/2014 6:55:35 PM

mortality is roughly 5%. Graft failure commonly results from sepsis. Other
common complications include hepatic artery thrombosis (5%), biliary
stricture (10%), biliary leak (<5%), and primary nonfunction (1%).
Cirrhosis portends a high risk for surgical and anesthetic complications,
including death. Careful assessment of the status of a patient’s liver dis-
ease is warranted prior to any elective procedure. Even with preoperative
risk stratification, cirrhosis may not be discovered until the time of sur-
gery. Most studies evaluating surgical risk in cirrhosis exclude Child class
C patients, and the majority of included cases are Child class A. This hetero-
geneity leads to poor generalizability of study results and, often, a misclas-
sification of surgical risk. As with any disease process, emergency surgery is
extremely high risk, with a high mortality rate. Estimates of mortality rates
for surgery in cirrhotics range from 11% to 25% versus 1.1% in noncirrhotic
patients. In general, the postsurgical 30-day mortality rate in Child class A
is 10% to 15%, Child class B is 30% to 40%, and Child class C is 75% to 80%.
Despite advances in surgical and anesthetic technique, these numbers are
largely unchanged over several decades.
The cause of these high mortality rates in cirrhosis is related to four
main organ systems: circulatory, pulmonary, immune, and hematologic.
Cirrhotics have a hyperdynamic circulation, and decreased hepatic perfu-
sion and anesthetic administration can exacerbate this, causing increased
hypotension and hypoxemia. Hypoxia may be exacerbated by previously
mentioned pulmonary complications due to ascites, hepatic hydrothorax,
portopulmonary hypertension, or HPS. Cirrhotic patients are more suscep-
tible to bacterial infection and subsequent sepsis, as well as wound healing
complications. Additionally, thrombocytopenia and coagulopathy lead to
bleeding complications. Malnutrition is prevalent among cirrhotic patients
and can exacerbate the aforementioned complications. Given the baseline
fluid derangements in these patients, fluid management intra- and postop-
eratively may be quite challenging.
The CTP class is used most widely in the literature to classify severity of
liver disease and surgical risk. It is easy to calculate, but it can have interob-
server variability, as the category of ascites and encephalopathy is subjective in
nature. The MELD score is more objective and gives weights to each variable. It
has been proven to be a good predictor of 30-day postoperative mortality, dem-
onstrating a linear relationship to mortality. Both CTP and MELD score may
be used in practice when determining surgical risk and guiding clinical man-
agement of a cirrhotic patient. Certainly, risk is optimized when a cirrhotic
patient is treated at a transplant center with an intensive care unit available.
CONCLUSION
Cirrhosis and subsequent portal hypertension are serious, life-threatening
diseases that may go undetected for years and may eventually lead to HCC.
Eventually, patients will succumb to the disease without a liver transplant,
although there are various ways to optimize medical management as a
bridge to transplant. Portal hypertension can result from cirrhosis, hepatic
vein obstruction, or portal vein thrombosis, all of which have many poten-
tial causes and may not be related to liver dysfunction. Variceal bleeding
is a life-threatening complication of portal hypertension and may be the
presenting symptom of end-stage liver disease. Detection of the subtle signs
and symptoms on history and physical, as well as careful interpretation of
laboratory and diagnostic imaging, can lead to early identification of cir-
rhosis cessation of the inciting factors of liver damage.
002086379.INDD 138 7/14/2014 6:55:35 PM

EDITORIAL NOTE
The profound clinical importance of hepatic cirrhosis is impossible to under-
state. For example, contemporary 1-year mortality of cirrhotic patients after
a first episode of esophageal variceal hemorrhage or new onset of ascites
is 50%!
Many surgeons have experienced the disastrous consequences of oper-
ating on cirrhotic patients (even those with seemingly well-compensated
cirrhosis)—intraoperative and occasionally delayed postoperative bleeding,
development of postoperative ascites (often leaking through abdominal
incisions) with secondary bacterial peritonitis; hepatic insufficiency;
encephalopathy (commonly accompanied by aspiration/pneumonia); and
almost astonishingly high mortality, even after simple procedures. Cirrhotic
patients have no physiologic reserve and little margin for error. These
patients should be approached by an experienced surgical team.
Suggested Readings
Albano E. New concepts in the pathogenesis of alcoholic liver disease. Expert Rev
Gastroenterol Hepatol 2008;2:749–759.
Al-Gusafi SA, McNabb-Baltar J, Farag A. Clinical manifestations of portal hyperten-
sion. Int J Hepatol 2012:2012:203794.
Anthony PP, Ishak KG, Nayak NC, et al. The morphology of cirrhosis: definition,
nomenclature, and classification. Bull World Health Organ 1977;55:521–540.
Chedid A, Mendenhall CL, Gartside P, et al. Prognostic factors in alcoholic liver dis-
ease. VA Cooperative Study Group. Am J Gastroenterol 1991;86:210–216.
Csikesz NG, Nguyen LN, Tseng JF, et al. Nationwide volume and mortality after elec-
tive surgery in cirrhotic patients. J Am Coll Surg 2009;208:96–103.
Grattagliano I, Ubaldi E, Bonfrate L. Management of liver cirrhosis between primary
care and specialists. World J Gastroenterol 2011;17(18):2273–2282.
Kahn S, Tudur SC, Williamson P. Portosystemic shunts versus endoscopic therapy
for variceal rebleeding in patients with cirrhosis. Cochrane Database Syst Rev
2006(4);CD000553.
Nicoll A. Surgical risk in patients with cirrhosis. J Gastroenterol Hepatol 2012;27:
1569–1575.
Shaw JJ, Shah SA. Rising incidence of hepatocellular carcinoma in the USA: what does
it mean? Expert Rev Gastroenterol Hepatol 2011;5:365–370.
0002086379.INDD 139 7/14/2014 6:55:35 PM

12 Intraoperative Ultrasound
of the Liver, Bile Ducts,
and Pancreas
Nicholas J. Zyromski
INTRODUCTION
Intraoperative ultrasound (IOUS) plays a critical role in HPB surgery. In
general, IOUS has four purposes: (1) acquisition of new information (such
as identifying occult metastatic disease), (2) complementing preoperative
and intraoperative radiologic imaging (cross-sectional imaging, intraop-
erative cholangiography), (3) guiding surgical procedures (biopsy, abla-
tion), and (4) confirming operation completion (clearance of stones from
the common bile duct, confirming vascular flow following reconstruction).
Structured ultrasound instruction has been developed by the
American College of Surgeons, ranging from trauma uses ( focused abdomi-
nal sonography for trauma—FAST) to endocrine (thyroid/parathyroid),
breast, vascular, and other applications. Similarly, leaders in HPB surgery
education recognize IOUS as an essential part of HPB training and are cur-
rently working to formalize training objectives. The goal of this chapter is to
provide a review of ultrasound basics and a practical approach to IOUS of
the liver, biliary tree, and pancreas.
BASIC PHYSICS
Sound energy passes through matter at various frequencies. The human ear
perceives sound waves ranging in frequency from 20 to 20,000 hertz (Hz, cycles
per second). “Ultrasound” by definition includes sound waves above the level
of human hearing, that is to say greater than 20,000 Hz. Medical ultrasound
applications typically range between 1 and 15 million Hz (mega hertz or MHz).
Sound waves move through tissues at different speeds depending on
the tissue density. Some of the sound waves are reflected, while other sound
waves will attenuate at different rates depending on tissue density. The rela-
tive amount of sound wave reflection and attenuation is a function of the
tissue impedance.
Electrical current applied to crystals in the ultrasound probe gener-
ates sound waves at specific frequencies (typically a range or “bandwidth”
of frequencies, i.e., 4 to 9 MHz). These sound waves pass through tissue and
reflect from tissue. The probe “listens” for returning sound waves—in fact,
up to 99% of a probe’s time is spent “listening” as opposed to sending out
sound waves. These returning sound waves are then converted or trans-
duced back to electrical energy from which an image is generated. This sig-
nal transduction from electrical voltage to sound wave and back is known
as the piezoelectric effect.
BASIC ULTRASOUND MECHANICS—“KNOBOLOGY”

The appearance of an ultrasound console may be quite intimidating.
(Fig. 12.1 shows the console of a typical IOUS unit.) The next few paragraphs
140
002153082.INDD 140 7/14/2014 7:01:11 PM

Chapter 12 / Intraoperative Ultrasound of the Liver, Bile Ducts, and Pancreas 141
FIGURE 12.1 Console of Aloka ultrasound unit. Gain knob is on bottom (long
arrows); TGC slider pots are on top right (short arrows).
will break down this big picture to smaller, usable, practical segments.
Pattern recognition is important in IOUS.
Echogenicity refers to the ultrasound appearance of tissues and
structures. Anechoic structures appear to be without echoes; hypoechoic
structures have fewer echoes, and hyperechoic structures have brighter
echoes than surrounding structures. (Fig. 12.2 shows the gallbladder lumen
FIGURE 12.2 IOUS of the gallbladder (short

arrow, note hyperechoic stones) in a patient
with portal vein thrombosis and cavernous
transformation (darker veins, long arrow).
0002153082.INDD 141 7/14/2014 7:01:13 PM

that is anechoic or hypoechoic relative to the adjacent liver parenchyma, as

well as hyperechoic stones in the gallbladder lumen.)
Signal transmission: begins with coupling of the transducer and
the structure being imaged. Coupling may be generated by direct probe
placement on the tissue to be imaged (liver, pancreas, etc.) or via “stand-
off ” through an acoustic window such as the stomach/duodenum.
Flooding the operative field with saline and using this fluid as an acous-
tic window is a useful standoff technique with which to evaluate super-
ficial lesions.
Probe types: Transducer frequencies ranging from 5 to 10 MHz are use-
ful for IOUS. Many IOUS probes are set at 7.5 MHz. The sound penetration of
these frequencies is 6 to 10 cm, which is acceptable for most IOUS applica-
tions. Most IOUS probes are linear array; the crystals may be arranged in
a linear fashion (which generates rectangular images) or curved orienta-
tion (also called convex probes). Probes may be of the “finger” or “T” shape.
Each probe shape has advantages and limitations. The operator should try
as many probes as possible; most will gravitate to one probe type or another
(i.e., finger or T). Laparoscopic probes may be rigid or articulate in one or
two orientations; the latter are particularly helpful when scanning the pos-
terior liver sections (Fig. 12.3).
Orientation: Two basic planes are recognized in ultrasound—longi-
tudinal and transverse. The operator should orient the probe (particularly
the T probe) by pattern recognition—the left lateral section of the liver is
a useful landmark with the “point” of the left lateral section aiming to the
left upper corner of the screen (Fig. 12.4). The right kidney also serves as a
good landmark.
FIGURE 12.3 Articulating laparoscopic transducers provide more range of tissue coverage.
002153082.INDD 142 7/14/2014 7:01:15 PM

FIGURE 12.4 Orientation relative to left lateral liver section. A. Correct orienta-
tion, edge of the liver at top right of screen (arrow). B. Backward orientation.
Depth: can be adjusted in many ultrasound units (Fig. 12.1). Users

should remember that image “magnification” typically sacrifices some
clarity.
Image refinement: Once orientation and ideal scanning depth have
been established, the image is refined by adjusting the gain and time gain
compensation (TGC). Adjusting gain amplifies the returning echo signal
(makes the image on the screen brighter). The gain should be set initially
to make structures known to be anechoic (such as blood vessels) black. Too
little gain results in nonvisualization of some structures; too much gain
results in image artifact. Sound waves attenuate as they pass through tis-
sue. The TGC controls are used to “fine tune” the depth of focus of IOUS
images, using the most sensitive area of the sound wave (the focal region of
transition from near field to far field) to interrogate the structure of interest.
Scanning techniques: systematic scanning is undertaken using four
basic probe movements—sliding, rotating, rocking, and tilting. The novice
ultrasonographer’s initial impulse is to slide the probe; however, a remark-
able amount of information can be learned simply by rocking and tilting
the probe from one discrete position. The operator must be aware of the
effect probe pressure exerts on the underlying tissue. Too much pressure
may compress vascular structures (Fig. 12.5) or obscure lesions. On the
other hand, deliberate compression (“compression scanning”) is a useful
technique for defining some structures. For example, in the porta hepatis,
the portal vein will compress, while the hepatic artery will not compress.
Special techniques: such as standoff, acoustic window use, or intra-
vascular contrast-enhanced ultrasound may be applied to appropriate
situations.
0002153082.INDD 143 7/14/2014 7:01:16 PM

FIGURE 12.5 Compression. A. Liver without compression. Note hepatic vein (short arrow),
portal vein branch (long arrow, note hyperechoic edges from extension of Glisson capsule
around the portal triad), and hyperechoic ligamentum venosum (dashed arrow). B. Probe is
in the same position, though with compression. Note ligamentum venosum (dashed arrow)
as anatomic reference; the vascular structures have been obliterated by the compression.
002153082.INDD 144 7/14/2014 7:01:17 PM

Mode: most ultrasound images are in gray scale “B” mode (bright-
ness modulation). Doppler mode may be used to evaluate vascular flow.
In Doppler mode, spectral analysis is plotted against time. Color flow uses
computer transformation of Doppler direction and velocity. By convention,
flow toward the transducer is red and flow away from the transducer is
blue. Velocity is typically shown in shades—darker shades indicate reduced
velocity, and brighter hues representing more rapid velocity.
Annotation and storage: Images may be stored either as static images
or as “cine” loop recording real-time scans.
Biopsy/ablation: biopsy or ablation may be directed precisely by ultra-
sound. It is critical to confirm needle position in two planes. A helpful tech-
nique is to apply color flow: moving the biopsy needle or ablation probe
under color flow visualization produces color motion marking.
Table 12.1 illustrates basic approach to IOUS scanning.
INTRAOPERATIVE ULTRASOUND OF THE LIVER

Intraoperative liver ultrasound is used to evaluate known lesions, identify
new lesions, and guide procedures (resection, ablation, biopsy). This guid-
ance also includes precisely localizing a lesion’s location relative to intra-
hepatic vascular structures, which has obvious therapeutic implications.
Despite advances in cross-sectional imaging, IOUS remains the most accu-
rate way to detect liver lesions; as many as 10% of patients will have unan-
ticipated additional liver lesions found at the time of IOUS evaluation. This
situation is particularly true for neuroendocrine tumors, many of which
have numerous occult intrahepatic metastases.
Intrahepatic metastases have a discrete acoustic character—they
may be isoechoic, hyperechoic, or hypoechoic relative to the surrounding
liver parenchyma (Fig. 12.6). Multiple metastases in the same patient nearly
always have similar echogenicity. Therefore, the best approach to a patient
with known hepatic metastases is to first identify a known lesion (i.e., one
that is palpable or easily seen on cross-sectional imaging) to determine
echogenicity. After identifying the echogenic characteristics of the known
lesion, the entire hepatic parenchyma should be scanned systematically,
keeping in mind the lesions’ character. Isoechoic lesions obviously are the
most challenging to detect intraoperatively.
In general, the approach to liver IOUS involves documenting vascular
inflow and outflow, followed by systematic evaluation of the parenchyma
with particular attention paid to target lesions’ relationship to vascular
TABLE
12.1 Basic Approach to IOUS Scanninga
1. Orientation—relative to known patterns such as left lateral liver section

2. Depth adjustment—will sacrifice resolution for magnification
3. Image refinement—gain and TGC
4. Systematic scanning
a. Movements—sliding, rotating, rocking, tilting
b. Compression scanning
c. Special techniques—standoff, acoustic windows
d. Mode adjustment—duplex, color flow
e. Biopsy/ablation (if indicated)
5. Annotation and storage
6. Completion scan
a
IOUS should be performed both prior to and after any dissection.
0002153082.INDD 145 7/14/2014 7:01:17 PM

FIGURE 12.6 Metastatic cholangiocarcinoma to the

liver (arrow). The lesion is hypoechoic to the surround-
ing hepatic parenchyma.
structures. Mobilization of the liver permits more accurate ultrasono-

graphic examination, though also creates superficial artifact. Therefore,
liver IOUS should be performed both before and after mobilizing dissection.
Systematic Approach
Again, IOUS of the liver consists of inflow, outflow, and parenchymal evalu-
ation. Pattern recognition is important: an example is the fact that hepatic
veins have more acute angles of articulation than portal veins, which join at
more of a right angle. Particularly important to inflow is evaluation of known
tumor relationship to vascular structures. An example of this concept is
a caudate tumor’s relationship to the inferior vena cava and portal vein.
Outflow imaging documents hepatic vein drainage to the suprahepatic vena
cava as well as documentation of accessory right hepatic vein. Although
this anatomy is ideally known from preoperative imaging, the actual spa-
tial relationship to the vena cava is well visualized with the ultrasound. The
relationship of central tumors to the intrahepatic vasculature (particularly
the middle hepatic vein) is similarly quite important. Intravascular tumor
thrombus may be diagnosed by IOUS; this finding may dramatically alter
operative approach.
Hepatic parenchymal pathology such as cirrhosis, steatosis, and post-
chemotherapy changes significantly affects ultrasound characteristics. The
hyperechogenicity of fatty liver or cirrhotic liver limits IOUS resolution.
Evaluation of the parenchyma should be performed in a systematic fashion,
that is, “lawn mowing” or segmental approach. No one perfect method for
parenchymal scanning exists; the operator should become familiar with an
approach and apply this approach consistently. As mentioned above, intra-
hepatic metastases may be hyper-, hypo-, or isoechoic to the underlying
002153082.INDD 146 7/14/2014 7:01:18 PM

parenchyma; identifying the echogenicity of a known lesion early facilitates

systematic interrogation of the hepatic parenchyma. Parenchyma should
be scanned slowly. The operator should be cautious of superficial lesions
that may not be completely captured by ultrasound without using standoff
techniques. The operator should also be aware of compression, too much
of which may obscure intrahepatic lesions. With increasing experience,
the ultrasound operator will become more comfortable adjusting TGC to
obtain more precise visualization of deep liver lesions. Perihepatic lymph
nodes particularly the portal nodes, celiac nodes, and retropancreatic
(aortocaval) nodes should also be evaluated at this point. Finally, postre-
section, evaluating flow in vascular structures with Doppler/flow mode,
is particularly useful after vascular reconstruction or to document portal
flow following a Pringle maneuver.
Pearls
■■ Interoperative ultrasound of the liver may be limited in the setting of
hepatic parenchymal pathology such as cirrhosis and steatosis.
■■ Do not “directly observe” intrahepatic radiofrequency ablation; the heat
may damage the sensitive ultrasound crystals as they contact the liver
surface.
■■ For ultrasound-guided needle localization (radiofrequency ablation or
biopsy), the angle of needle placement may be challenging, and experi-
ence is helpful in determining this approach. Needle placement directly
adjacent to the ultrasound probe may not be the best position.
■■ Validating probe position in two planes is mandatory prior to ablation.
This maneuver must be performed prior to ablation to avoid artifact.
Also, remember that ablation defects are not spherical.
■■ Simple cysts of the liver are easily characterized; these lesions are
anechoic with posterior shadow (Fig. 12.7).
FIGURE 12.7 Classic features of simple

liver cyst: anechoic cyst with posterior
shadowing (arrows).
0002153082.INDD 147 7/14/2014 7:01:19 PM

■■ Hemangioma of the liver may be confirmed by documenting flow through

the lesion (although this diagnosis is almost always secure preoperatively
with adequate IV contrast-enhanced cross-sectional imaging).
■■ The ultrasound character of focal nodular hyperplasia and hepatic ade-
noma is very similar—these heterogeneous lesions are not possible to
distinguish by IOUS.
■■ The posterior section and segment VII in particular is most challenging
to visualize with the laparoscopic ultrasound probe. A two-way articu-
lating probe is quite helpful evaluating these areas. Mobilizing the fal-
ciform and coronary ligaments also helps expose the posterior section
for IOUS.
■■ It is important to complete laparoscopic IOUS through multiple port
sites.
■■ Scoring the liver capsule with electrocautery provides a useful artifact
that helps define the intrahepatic plane of resection.
INTRAOPERATIVE ULTRASOUND OF THE BILIARY SYSTEM

IOUS of the biliary system is most commonly used to visualize intrahe-
patic and extrahepatic stones as well as to evaluate the relationship of
biliary tumors to intra- and extrahepatic vascular structures. Less com-
monly, the diagnosis of biliary cystadenoma may be confirmed by visual-
izing septa and/or rarely a direct communication between cyst and the
biliary tree. The later is usually seen more clearly with magnetic resonance
cholangiography.
Systematic Approach
It is helpful to use an acoustic window (i.e., viewing through hepatic seg-
ment III or IV) to visualize the extrahepatic biliary tree. Similarly, either the
stomach or the duodenal bulb provides a good acoustic window to visual-
ize the intrapancreatic common bile duct. Of note, no Kocher maneuver
is necessary to image the intrapancreatic bile duct. If Kocher maneuver is
planned, ultrasonography should be performed both before and after dis-
rupting these tissue planes. The quality of ultrasound images decreases sig-
nificantly after dissection.
The course of the biliary tree should be defined. Particular atten-
tion should be paid to the aberrant biliary anatomy such as a low-lying
right posterior sectoral duct. The relationship of the biliary tree to vas-
cular structures (particularly the right hepatic artery) should be defined.
Figure 12.8 shows aberrant right hepatic artery coursing anterior to the
common bile duct. The presence of shadowing defects (e.g., intraductal
stones) should be sought. Three classic sonographic criteria for stones are
(1) presence of hyperechoic focus, (2) posterior shadowing, and (3) mobil-
ity (Fig. 12.9). Most stones are mobile with movement of the gallbladder
or bile duct—this distinguishes them from polyps or tumors, which are
fixed. Gallbladder polyps may not shadow; generally, neoplastic polyps
are less echogenic than cholesterol polyps. Infiltrating tumors may show
irregular wall thickening; attention should be paid to the interface of the
gallbladder wall with the hepatic parenchyma. Tumors of the gallbladder
fundus or extrahepatic biliary tree should be evaluated relative to their
relationship with adjacent vascular structures—portal veins and hepatic
arteries.
Although some skilled operators advocate IOUS as the sole modal-
ity to diagnose bile duct stones, it is important to realize that the learn-
ing curve for biliary IOUS is relatively steep. Many biliary surgeons
feel comfortable confirming ultrasound findings with intraoperative
cholangiography.
002153082.INDD 148 7/14/2014 7:01:19 PM

FIGURE 12.8 Transhepatic acoustic window visualizing extrahe-

patic biliary tree. Aberrant right hepatic artery (arrows) passing
anterior to dilated common bile duct (star).
Pearls
■■ IOUS is helpful in identifying the course of an aberrant arterial structure
(i.e., replaced right hepatic artery, right hepatic artery anterior to the
common hepatic duct).
■■ IOUS of the biliary system may actually be most useful in laparoscopic
surgery, where decreased tactile sensation is available.
■■ IOUS is quite useful to evaluate the intrahepatic and intrapancreatic
extension of choledochal cysts.
■■ With practice, ultrasonography allows identification of the common bile
duct and main pancreatic duct junction in the pancreatic head, that is,
an abnormal biliopancreatic junction.
■■ IOUS is useful to localize the papilla for “targeted” (limited) duodenot-
omy for papillary exploration, such as sphincteroplasty, ampullectomy,
transsphincteric exploration.
■■ IOUS is helpful to identify enlarged and possibly pathologic involved second
echelon lymph nodes in patients with cholangiocarcinoma and gallblad-
der cancer (i.e., celiac, retropancreatic, aortocaval nodes). Many authorities
consider involvement of these nodes to be a contraindication to resection.
■■ In gallbladder cancer, ultrasound is important in evaluating tumor rela-
tionship to the portal vein and the hepatic artery.
■■ In Mirizzi syndrome, ultrasound is quite useful for evaluating the biliary
tree in relationship with the surrounding vascular structures and posi-
tion of the stone.
■■ Pressure from the ultrasound probe in the porta hepatis will compress
the portal vein but not the hepatic artery.
■■ Intrahepatic bile ducts are typically not visible on ultrasound scanning.
Therefore, any intrahepatic biliary dilation visible on IOUS suggests
biliary pathology.
0002153082.INDD 149 7/14/2014 7:01:20 PM

FIGURE 12.9 Laparoscopic IOUS of common bile duct stone

(arrow) shown in transverse (A) and sagittal (B) orientation.
The stone is hyperechoic with posterior shadowing.
002153082.INDD 150 7/14/2014 7:01:21 PM

INTRAOPERATIVE ULTRASOUND OF THE PANCREAS

Ultrasound is a remarkably useful but generally underused modality with
which to evaluate pancreatic pathology. Similar to most tools, the more one
uses ultrasound for pancreatic evaluation, the more comfortable the opera-
tor becomes with the approach and the more they appreciate the broad
variety of applications of this modality.
Systematic Approach
Ultrasound of the pancreas should be performed both using an acoustic
window such as the stomach or duodenum as well as by placing the probe
directly on the pancreatic parenchyma to appreciate improved resolution.
Systematic approach to the pancreatic ultrasound involves evaluation of the
parenchyma, pancreatic duct, vascular relationships, and the target lesion
(i.e., tumor, stone, dominant stricture). The pancreas relationship to adjacent
vascular structures such as the splenic vein/superior mesenteric vein conflu-
ence (Fig. 12.10) provides useful points of reference (i.e., pattern recognition).
Pancreatic Parenchyma
Normal pancreatic parenchyma is usually homogeneous, with a small,
anechoic or hypoechoic duct visible. The typical “salt and pepper” speckled
features of the pancreatic parenchyma is due to the hyperechogenic intra-
parenchymal fat. Pancreas echogenicity is usually similar to or greater than
that of the liver.
Features of chronic pancreatitis have been defined to include parenchy-
mal (hyperechoic foci, hyperechoic strands, hypoechoic lobules, and cysts)
and ductal (dilation, dilated side branches, main duct irregularity, hyperechoic
duct margins, and stones) criteria (Fig. 12.11). Table 12.2 outlines these criteria.
FIGURE 12.10 Transverse orientation at the pancreatic head. Superior mesenteric vein
(SMV) (in A) and SMV/splenic vein confluence (in B) are shown with long arrow ; pancreatic
duct at the genu (A) and body (B) is shown with short arrows.
0002153082.INDD 151 7/14/2014 7:01:22 PM

FIGURE 12.11 Chronic pancreatitis showing main

pancreatic duct stricture (short arrow) with dilation of
MPD upstream to the tail (long arrow).
Pancreatic Duct
The pancreatic duct should be evaluated for size, tortuosity, known (or
unappreciated), strictures, the presence of pancreatic divisum, and/or size
of accessory pancreatic duct, which has implications when planning trans-
duodenal sphincteroplasty. A minimally dilated or tortuous pancreatic
duct may be very challenging to open longitudinally if one is considering
a lateral pancreaticojejunostomy; direct IOUS guidance is quite help-
ful in this circumstance. Perhaps, the most important relationship of the
pancreatic duct is that of the main pancreatic duct to a cyst or neuroen-
docrine neoplasm if the operator is considering enucleation. Injury to the
main pancreatic duct results in a high-volume, morbid pancreatic fistula.
TABLE
12.2 Ultrasound Characteristics of Chronic Pancreatitis
Parenchymal
Hyperechoic foci
Hyperechoic strands
Hypoechoic lobules
Cyst
Ductal
Irregular duct contour
Visible side branches
Hyperechoic duct margin
Dilated main duct
Stones
002153082.INDD 152 7/14/2014 7:01:23 PM

The intrapancreatic portion of the common bile duct should be evaluated

routinely for size, intraluminal lesion, and relationship to pancreatic ducts.
Vascular Relationships
Vascular relationships include relation of the pancreas and target lesion to
the surrounding arterial and venous structures: portal, splenic, and superior
mesenteric veins, and celiac, hepatic, gastroduodenal, and superior mesen-
teric arteries. This relationship is unique to individual lesion/tumor posi-
tion. Often the splenic artery is intraparenchymal in the body and the tail of
the pancreas. Therefore, understanding splenic artery anatomy is important
when the operator is considering spleen preserving distal pancreatectomy
or lateral pancreaticojejunostomy (which should be extended to the tail).
Unusual anatomy such as a replaced right or left hepatic artery is important
and may be recognized with the IOUS. The presence and/or extent of portal
vein/superior mesenteric vein or splenic vein thrombosis can be identified
and again may have direct implications on operative planning.
Target
Finally, in the systematic approach to pancreas IOUS is evaluation of the
target lesion. Tumors should be evaluated for their relationship to adjacent
vasculature (i.e., superior mesenteric/portal veins, superior mesenteric
artery,) as well as to adjacent organ structures such as the stomach, adre-
nal, and colon in the case of pancreas tail tumors. Deliberate characteriza-
tion of tumor echogenic features will inform the operator’s future practice.
For example, pancreatic adenocarcinoma is typically hypoechoic; distal
cholangiocarcinoma may not obstruct the pancreatic duct (Fig. 12.12);
FIGURE 12.12 Longitudinal view of a metal biliary

stent passing through well-circumscribed, hypoechoic
distal cholangiocarcinoma (arrows). The pancreatic duct
in this patient was not dilated.
0002153082.INDD 153 7/14/2014 7:01:24 PM

FIGURE 12.13 Transverse view of a pancreas neuroendocrine

tumor is hyperechoic and well circumscribed, with hypoechoic
edges and faint posterior shadowing. Superior mesenteric vein
is shown with arrow.
and neuroendocrine neoplasms are typically hyperechoic and well-circum-

scribed relative to surrounding parenchyma (Fig. 12.13).
Dominant strictures in the pancreatic duct should be identified; in
addition, ultrasound provides a sensitive method to interrogate the entire
ductal system for other unexpected strictures, which may influence the
choice of operation, particularly in chronic pancreatitis patients.
Pancreatic cysts are easily identified and characterized by IOUS
(Fig. 12.14). Attention should be directed toward communication with the
main pancreatic duct, wall irregularity, or mural nodularity. In addition, the
relationship of the cyst to surrounding structures including the main pan-
creatic duct is important, especially if considering enucleation.
IOUS is the most accurate method with which to characterize peri-
pancreatic inflammatory collections in the setting of necrotizing pancre-
atitis (Fig. 12.15). The volume of solid necrosis has major implication in
treatment planning.
Pearls
■■ Fine tuning TGC is particularly useful when evaluating the vascular/
tumor interface.
■■ Knowledge of the relationship between cysts or islet cell tumors and the
main pancreatic duct is critical when considering enucleation.
■■ In laparoscopic IOUS evaluation of the pancreas, port placement is impor-
tant. The pancreas should be scanned from at least two separate ports.
The operator should also be aware that probe orientation and the result-
ing image will change significantly with different fulcrum of laparoscopy.
002153082.INDD 154 7/14/2014 7:01:25 PM

FIGURE 12.14 Transverse orientation, laparoscopic IOUS of

pancreatic tail mucinous cystic neoplasm (MCN; arrow). Note
thick walls and internal septae. Note also relationship to the
left kidney (star).
■■ Evaluation of the pancreatic duct is important when considering open-

ing a small pancreatic duct for lateral pancreaticojejunostomy.
■■ It bears repeating that ultrasound is the most accurate way to iden-
tify solid versus fluid peripancreatic necrotic debris in the setting of
necrotizing pancreatitis.
FIGURE 12.15 Transgastric view (transverse orientation) of peripancreatic

collection in necrotizing pancreatitis patient. This collection has fluid (stars)
and solid (arrows) components.
0002153082.INDD 155 7/14/2014 7:01:26 PM

FIGURE 12.16 Laparoscopic ultrasound scanning from several different ports.

(From Machi J, Staren E, eds. Ultrasound for surgeons, 2nd ed. Philadelphia,
PA: Lippincott Williams & Wilkins, 2005.)
INTRAOPERATIVE LAPAROSCOPIC ULTRASOUND

Laparoscopic ultrasound follows the same principles as other IOUS tech-
niques. The major difference is limited probe motion based on probe rigid-
ity and the fulcrum of the probe interface through the abdominal wall
(Fig. 12.16). Laparoscopic probes that articulate in one and two directions
provide better tissue coverage than rigid probes (Fig. 12.3). Laparoscopic
ultrasound is technically and intellectually more demanding than open
IOUS, with an attendant longer learning curve. Laparoscopic IOUS should
be performed through multiple ports (i.e., umbilical, subxiphoid, lateral).
The operator should concentrate on understanding image patterns that
may be markedly different from traditional (open) IOUS patterns.
CONCLUSION
In summary, intraoperative ultrasonography is an extremely valuable addi-
tion to the HPB surgeon’s armamentarium. Performing ultrasound on every
case is helpful to understand both normal parenchymal and anatomic
relationships as well as becoming familiar with pathology and unusual
relationships.
Suggested Readings
Choti MA, Kaloma F, de Oliveira ML, et al. Patient variability in intraoperative
ultrasonographic characteristics of colorectal liver metastases. Arch Surg
2008;143(1):29–34.
Connor S, Barron E, Wigmore SJ, et al. The utility of laparoscopic assessment in the
preoperative staging of suspected hilar cholangiocarcinoma. J Gastrointest Surg
2005;9(4):476–480.
Hagopian EJ, Manchi J, eds. Abdominal ultrasound for surgeons. New York, NY:
Springer, 2014.
002153082.INDD 156 7/14/2014 7:01:27 PM

13 Energy Devices for
Parenchymal Transection in
Liver Surgery
Prejesh Philips and Robert C.G. Martin II
BACKGROUND
The first well-documented reports of deliberate resection of liver tumors
come from A. Luis in 1886 and Carl von Langenbuch in 1887, both of
which were marred by postoperative bleeding. Control of the hepatic
bleeding was accomplished with “well-placed sutures and well-timed
prayers.” William Williams Keen was credited with first liver resection
in America and also described the “finger fracture” technique and indi-
vidual cauterization. In1888, Hugo Rex and, in 1897, James Cantlie from
Liverpool rediscovered the work of Glisson from 1654 and described the
avascular plane through the gallbladder bed toward the vena cava, now
known as the Rex-Cantlie line. This was followed by description of the
Pringle maneuver, and finally the seminal work of Claude Couinaud, in
1954, on the segmental architecture of the liver, led to the blossoming of
modern-day hepatic surgery. The finger fracture technique was initially
described and was followed later by the clamp–crushing technique or
Kelly-clysis, which has since become the reference parenchymal transec-
tion method. Following these anatomical principles and improvement in
anesthesia, surgeons could attain anatomical resection with less blood
loss, but the problem of bleeding parenchyma, erratic vasculature, high
mortality, and morbidity persisted. At the same time, it has been shown
clearly that the use of blood transfusions in liver surgery is associated with
worse outcomes. In an attempt to decrease blood loss and improve tran-
section speed and quality, various hemostatic assist energy devices have
been designed.
Although great advances have been made in technology, the ideal
hemostatic energy device and mechanism is a topic of much debate. The
ideal coagulation mechanism would include ease of use, minimal lateral
energy damage, the ability to cut as well as coagulate, control of tissue
temperature (so that charring is avoided), reliability, noninterference with
medical devices, and the versatility to work on a variety of tissues. We will
individually examine the commonly used energy sources, review current
literature, and compare their strengths, weaknesses, and uses.
ELECTROSURGERY
Thermal and chemical cauteries were the first tools for hemostasis. This
technique evolved over time to the use of electrocautery or more accurately
electrosurgery, since it involves the use of an alternating current in which
the patient is part of the circuit. The principle is generation of electricity
from the Electro-generator unit, delivered to a handheld device, which
then uses tissue impedance to produce heat. The complementary part of
the circuit, return electrode, or grounding pad takes the electrons back to
complete circuit.
157
0002086380.INDD 157 7/15/2014 12:36:02 PM

Monopolar high-frequency electrical energy is the oldest method used

to coagulate vessels. The first commercial electrosurgical device is credited
to William T. Bovie and used for the first time in 1926 in Boston. It uses a
generator to create high-frequency current (200 kHz) in order to prevent
nerve conduction and generate local thermal energy while at the same time
preventing remote electrical injury. Because of this, a local thermoablative
zone is created. There are different modes that monopolar electrocautery
can be used. “Cut” causes a high enough voltage (> 400 V peak to peak) to
create a vapor pocket and results in an incision. It typically uses a modu-
lated periodic sine waveform. When the system is operating in “coag mode,”
the voltage output is usually lower than in the cut mode and less power is
delivered. Typically, sine wave is turned on and off in a rapid succession.
This therefore generates less heat, and a vapor pocket is not generated.
The overall effect is a slower heating process that causes tissue to coagu-
late and smaller vessels are destroyed and sealed, stopping capillary and
small arterial bleeding. In simple coagulation/cutting mode machines, the
lower-voltage cycle typical of coagulation mode is usually heard as a lower
frequency.
Monopolar electrocautery is versatile, almost universal in availability,
and cheap, acts quickly, and is relatively safe to use. It can be used in a vari-
ety of settings and can effectively help control bleeding from small vessels
and capillaries.
Limitation of Electrosurgery
Patients with electrical implants such as cardiac pacemakers require spe-
cial precautions, especially when using monopolar devices. Conventional
monopolar electrosurgery should not be used to control larger vascular
structures, around vascular compromised tissue, bowel, and other visceral
structures such as the bile duct and ureter. The limitations of monopolar
electrocautery include poor efficacy with vessels greater than 3 mm, a wide
collateral zone of thermal injury, and lack of an effective feedback mecha-
nism except visual approximation.
Complications of Electrosurgery
Electrosurgery-related complications are relatively uncommon, occurring
in 2 to 5/1,000 procedures. Lack of experience and knowledge of the device’s
mechanism of action is the major contributing factor with approximately
60 procedures being the inflection point for injuries from electrothermal
devices in laparoscopy.
The mechanism of electrosurgical injury includes direct injury, lat-
eral thermal spread, direct coupling, capacitive coupling, and insulation
failure. Direct coupling results from inadvertent contact of two noninsu-
lated instruments, and the current flows from the primary to the second-
ary instrument. Severe injury may result from a second conductor if it is in
contact with sensitive structures. Capacitive coupling is alternating current
flowing through the primary electrosurgical device inducing unintended
stray current in any conductor (such as metal trocars of liver retractors) in
close proximity with the monopolar instrument.
Hollow viscus injury is uncommon but is often missed and can have
fatal consequences. The nature of the thermal injury is such that intraop-
eratively, the injury may be masked and will manifest usually about 4 to
10 days postoperatively, depending upon the severity of the coagulation
necrosis. In the context of liver surgery, electrocautery has caused duodenal
injury and late bile duct strictures. These injuries can be avoided by mini-
mizing the use of electrocautery in extrahepatic pedicle control techniques.
While assessing the efficacy of the coagulation, attention should be paid
002086380.INDD 158 7/15/2014 12:36:02 PM

Chapter 13 / Energy Devices for Parenchymal Transection in Liver Surgery 159
to the whitening of tissue surrounding the tip of the electrosurgical instru-

ment and formation of bubbles. Whitening suggests adequate hemostasis
and bubbles represent water vapor; thus, the tissue is desiccated, and the
bubbles disappear, marking adequacy. Other visceral and vascular injuries
from electrocautery have been reported. Severe skin burns caused by a par-
tially detached grounding pad/return electrode can occur, but most mod-
ern electrosurgical monopolar devices have a return electrode monitoring
system to prevent this problem. Use of lowest-effective power settings,
intermittent activation, and direct visual feedback for adequacy are good
practices for the prevention of injury. With bipolar electrosurgery, termina-
tion at the end of the vapor phase and alternating between desiccation and
incision are a good practice.
Our preference is to use the monopolar electrosurgical device to gain
access to the abdomen and use it for mobilization of the right hepatic lobe
off the triangular ligaments. We also use it to “score” the liver capsule while
delineating resection lines. This demarcation defining the line of resection
can be seen with high-quality ultrasound imaging and helps to ensure an
adequate margin prior to resection. This line of demarcation for resection
is even more important when performing laparoscopic resections given the
limited visualization and lack of tactile feedback during resection.
Conventional Bipolar Electrosurgery

In bipolar electrosurgery, both the active electrode and return electrode
functions are performed at the site of surgery. The two tines of the forceps
perform the active and return electrode functions. Only the tissue grasped
is included in the electrical circuit. Because the return function is per-
formed by one tine of the forceps, no patient return electrode is needed.
Most bipolar units use a low-voltage waveform that achieves hemostasis
without excessive charring. For the same reasons, bipolar electrosurgical
devices are less effective for cutting tissue since adequate vaporization is
difficult to achieve. Effective hemostasis can be achieved by coapting and
thermally welding the blood vessels. Conventional bipolar is widely used in
neurosurgery and gynecologic surgery.
Bipolar electrosurgery technology is very safe because of limited
tissue conduction. Use of lower voltages in a smaller circuit leads to less
lateral thermal spread and injury. Bipolar energy is also safer to use when
there is a question about the safety of using more powerful monopolar elec-
trosurgical units (i.e., implantable medical electromagnetic devices such as
pacemaker, defibrillator, pumps). There are some disadvantages to the use
of bipolar: Bipolar cannot spark to tissue, and the low voltages make it less
effective on large vessels.
BIPOLAR VESSEL-SEALING DEVICES

Ligasure
Vessels sealing devices stemmed from refinement of bipolar electrosurgery
technology. The bipolar vessel-sealing system (LigaSure) applies a precise
amount of bipolar energy and pressure to fuse collagen and elastin within
the vessel walls. This fusion is done using a feedback-controlled energy
delivery through a generator that senses adequate coaptation and coagu-
lation. In addition, the device minimizes lateral thermal spread. The seal
with this device can withstand three times the normal systolic pressure and
seals vessels up to 7 mm. The sealing is achieved with minimal sticking and
charring; thermal spread to adjacent tissues is approximately 2 mm. Vessel
sealing has steadily gained acceptance in various surgical disciplines due to
its ease of use and ability to seal relatively large vessels. To underscore this
0002086380.INDD 159 7/15/2014 12:36:02 PM

point, vessel sealing has been successfully used on the splenic hilum. The
disadvantages include the need to fully engage and lock the clamps before
delivery of energy can occur and the nonuniform compression it delivers
over the length of the prongs. This technology is fast growing with propo-
nents arguing the ease of use, fast sealing, and relative cost–benefit versus
other devices. It is also one of the few devices with a cost–benefit analysis
in its favor in published literature.
Ligasure has fast been gaining ground in laparoscopic surgery and
has been used extensively in colorectal and bariatric procedures. A sys-
tematic review of six randomized controlled trials of hemostatic devices
in colectomy was favorable for the use of Ligasure over conventional elec-
trosurgery but did not demonstrate clear superiority over other ultrasonic
vessel-sealing devices. A Japanese randomized control trial (RCT) demon-
strated Ligasure’s superiority with respect to blood loss, transection times,
and lower number of ties required. They also showed a cost–benefit in using
Ligasure over conventional techniques. This was however followed shortly
later by another Japanese trial showing no benefit over the crush–clamp
technique. Another RCT, which showed superiority of the clamp–crush
technique, was conducted under inflow control and that needs to be taken
into consideration in the era of minimally invasive surgery where inflow
control is the exception rather than the norm.
EnSeal
Enseal is another vessel-sealing system that uses bipolar electrosurgical
principles. EnSeal delivers vessel sealing by using a combination of compres-
sion and thermocoagulative energy to ensure hemostasis. It is capable of
achieving seal strengths up to seven times the normal systolic pressures on
vessels up to 7 mm with a typical thermal spread of approximately 1 mm. It
differs from Ligasure in its compression mechanism, which applies uniform
pressure along the full length of the instrument jaw, achieving compression
forces similar to those of a linear stapler of up to 7,800 psi. Compression is
combined with controlled energy delivery utilizing Nanopolar thermostats
to reach collagen denaturation temperatures in seconds: temperatures
maintained at approximately 100°C throughout the power delivery cycle.
The device also has a simultaneous cutting mechanism to allow one-step
sealing and transection of vessels and soft tissues. The big cost advantage
is the ability to use conventional bipolar generators for its use, thereby
obviating a large new capital investment for generators. EnSeal also has
a large jaw width and produces minimal smoke (or steam). The dynamic
dual mechanism mandates active operator effort and cognizance and has a
modest learning curve as compared to other lock-fire devices.
This device is about 3 cm in length at the functional portion of the
head, which is comprised of jaws. The instrument head is designed to
facilitate atraumatic dissection. In a comparative porcine model, Enseal
devices were reported to achieve higher burst pressures with a sub-mm lat-
eral thermal spread. Another ex vivo study reported longer sealing times
and variable burst pressures with Enseal as compared to Ligasure device.
There have been few clinical studies of Enseal in published literature and
no randomized controlled trials. Initial studies demonstrated the safety
profile of this device. A single-center study on laparoscopic liver resection
using the EnSeal versus ultrasonic dissector showed shorter parenchymal
transection time. However, head-to-head comparisons with conventional
clamp–crush are not available in current literature. Despite this paucity
of clinical data, its performance in preclinical studies, ease of use, and
ability to simultaneously seal and divide large vessels without significant
lateral thermal spread have made Enseal popular among many centers.
002086380.INDD 160 7/15/2014 12:36:02 PM

Inherent limitations of the device remain the need for the operating sur-

geon to feel when the tissue is desiccated–sealed with a gradual compres-
sion of the device. The device differs from other devices as its mechanism of
action is initiated prior to complete closure of the device giving the operat-
ing surgeon feedback on the type of tissue that is being transected.
Salient Dissecting Sealer (Formerly TissueLink Monopolar)

The TissueLink floating ball is a saline-cooled, high-density monopolar
radiofrequency probe used for dissection and coagulation. Saline irrigation
at the tip maintains the surface temperature below 100°C, preventing char-
ring and tissue “sticking,” reducing smoke, and decreasing lateral thermal
spread. The saline facilitates energy transfer between the device–tissue
interface, maintaining contact with the hepatic tissue even when the tip is
moving and evenly dispersing thermal energy. The mechanism of sealing is
primarily by heat and shrinking, and it provides a clean transection bed and
can seal vessels up to 3 mm in size. It will isolate larger vessels, which can
then be controlled by other techniques.
TissueLink is useful in transecting cirrhotic livers and in resecting
lesions in close proximity to major vessels. Theoretically, the saline cooling
prevents lateral thermal spread, thus making it safer around biliary and
larger vascular structures. It is however expensive and slow. It has been
used in many centers in combination with ultrasonic dissection device.
Various nonrandomized studies showed efficacy in cirrhotic and noncir-
rhotic liver transection with respect to blood loss and postoperative liver
dysfunction, albeit with an increase in operative time. It has been used in
living donor hepatic transection showing superiority in one study over
CUSA and bipolar electrocautery. One randomized study, using water-jet,
ultrasonic dissector and aspirator, and salient dissecting sealer, found that
the dissecting sealer group used less hemostatic agents but was slower
and more expensive than Hydrojet and CUSA. Randomized trials however
have showed no significant benefit over the crush–clamp technique. The
benefits of this device remain its hemostatic potential and action, while
not requiring any type of compression. This can be very helpful when per-
forming nonanatomic resections in cirrhotic patients when the hepatic
parenchyma is more fibrotic that the portal inflow or hepatic outflow and
thus a crushing technique can lead to more bleeding. The limitations of
the device remain its methodical method of action and limited efficiency
when used alone.
PlasmaKinetic Tissue Management System

This system delivers pulsed bipolar energy through the instrument to the
tissue, allowing intermittent tissue cooling, which limits lateral thermal
spread and tissue sticking. The generator detects the optimal settings for
the specific instrument, and there are visual and audible tissue impedance
indicators. The system has two different modes: the vapor pulse coagulation
mode and the PlasmaKinetic tissue-cutting mode. In the vapor pulse mode,
high energy is delivered to grasped tissue, creating vapor zones. The current
then travels around the high-impedance vapor zones, following the path of
least resistance. The vapor zones subsequently collapse, and with each new
energy pulse, more and more tissue between the instrument jaws is coagu-
lated, ultimately resulting in uniform coagulation of tissue. It is capable of
sealing vessels 5 to 6 mm in diameter and has burst pressure up to three
times the systolic blood pressure with half of the lateral thermal spread
compared to conventional bipolar devices. In a recent study, PlasmaKinetic
was used for hepatic resections in 51 cirrhotic and noncirrhotic livers, with
acceptable morbidity and transection speeds. The authors point to the
0002086380.INDD 161 7/15/2014 12:36:02 PM

s tatistical similarity between the outcomes of the cirrhotics and noncir-

rhotic patients as evidence that it is efficacious in cirrhotic liver. There are
various effector instruments including hook, needle, and dissecting and
cutting forceps.
ARGON BEAM COAGULATORS

The argon beam coagulator is a noncontact, monopolar electrocoagulation
device that transmits radiofrequency electrical energy from a handheld
electrode across a jet of argon gas. The argon gas jet clears the field of pooled
blood and evenly distributes electrical energy to the target tissue. Argon
photocoagulation is used very commonly to control bleeding from the cut
parenchymal edge. It allows for rapid and diffuse superficial coagulation
at the plane of transection. It is usually used in conjunction with another
dissecting technique such as Kelly-clysis and CUSA. It cannot be used to
control large vascular and biliary structures, but with its low penetrance
contact coagulation, it is extremely effective in surface hemostasis. Its use
is almost ubiquitous in transplant and cirrhotic settings. There have been
various reports of bowel injury with its use when used by untrained hands
or because of a lack of understanding of the method of action of the device.
There remains a myth regarding the potential risk of venous gas embolism,
which can occur in any open case when the central venous pressure (CVP)
is low and/or inadvertent injury to the hepatic outflow structure occurs.
We have successfully utilized argon beam coagulation following over 300
laparoscopic cases and have not had a single venous embolism event, and
thus, we believe this risk is unproven when good surgical transection tech-
nique is utilized.
ABLATIVE TECHNOLOGIES
Radiofrequency-Assisted Parenchymal Transection (RF-PT): Habib 4 ×
Radiofrequency (RF) devices have been used for more than a decade to
thermoablate nonresectable hepatic lesions. The Habib was an extension
of this technology. It was one of the first devices to create a bloodless field
at the purported site of transection using thermocoagulative properties. It
uses RF probes, which can be inserted into the liver surface. This leads to a
plane of coagulative necrosis to be developed along the line of parenchymal
transection, with subsequent reduction of blood loss and transfusion. This
concept of precoagulation deep in the parenchyma of the liver before tran-
section allows nonanatomical resections. It can seal intrahepatic vessels
and bile ducts. The radiofrequency energy is typically applied in sequen-
tially overlapping segments to ensure adequate hemostasis. This technique,
however, is time consuming and costly. Despite the ability for nonanatomi-
cal resection, it does lead to a large area of hepatic eschar and more tissue
necrosis than other techniques.
In the RF-PT or radiofrequency-assisted parenchymal transection, the
Habib device 4 × has been used widely in various centers. Various nonran-
domized studies showed benefit especially in reduction of blood loss and
postoperative complications. However, there have been increasing reports
of liver failure and infectious complications in part due to the wide zone
of ablation that the RF device induces. In some randomized studies and
two meta-analyses, Habib was shown not to be superior to the crush–clamp
technique. Proponents argue that the use of RF-assisted device is especially
useful in cirrhotic livers where traditional transection methods can lead to
significant more blood loss. This claim has been backed up by at least one
randomized study that showed significantly lower blood loss and a trend
to decreased complications with the use of RF devices in cirrhotic livers.
002086380.INDD 162 7/15/2014 12:36:02 PM

It is useful in cirrhotic patients, but the risks of the large ablation zone lead-
ing to liver failure in these patients with marginal liver reserve need to be
taken into account. Additional challenges with this device for laparoscopic
use are the large trocar that is required and the lack of any type of transec-
tion feature leading to inefficient of movements during transection.
Microwave
Microwave energy technology has been used as an ablative therapy in
patients with nonresectable hepatic lesions. Similar to the RF-PT device,
it utilized the precoagulation principle. A microwave tissue coagulator
(Alfresa Ind., Osaka, Japan) generates a 2,450-MHz microwave, which is
transmitted to a monopolar type of needle electrode, called a microwave
scalpel, by way of a coaxial cable. This endoscopic surgical device has a
hand piece with needle-shaped monopolar applicators. Microwave emis-
sion is started after the liver tissue is punctured with a microwave probe or
scalpel. Molecular excitation by microwave causes thermal energy, which
results in zone of tissue coagulation around the microwave scalpel. The
coagulation region reaches a depth equal to the length of the scalpel. It can
be recognized as a superficial yellow-whitish coagulation zone approxi-
mately 1 cm in diameter.
Microwave precoagulation has been gaining favor among some sur-
geons. It has shown to be a safe and effective mode of precoagulation before
transection in a nonrandomized study with respect to bile leak, hospital
stay, and blood loss as compared to historical controls with other transec-
tion devices. It has a smaller ablation zone compared to radiofrequency
ablation. This technique however is time consuming and needs precoagu-
lation as a separate step. No randomized controlled trials have compared
microwave to other technologies or to the clamp–crush technique. Current
technologies have not been optimized for precoagulation and transection
with the use of microwave and thus remain limited to open resections.
DISSECTION-ONLY DEVICES
Ultrasonic Dissectors: (CUSA)
The Cavitron Ultrasonic Surgical Aspirator (CUSA) is a dissecting device
that uses ultrasonic frequencies to fragment tissue. In combination with
the aspiration of the dissected contents (i.e., hepatic parenchyma), CUSA
skeletonizes small vessels and bile ducts allowing control by clips, ligatures,
or other energy devices. Utilizing a hollow titanium tip that vibrates along
its longitudinal axis, fragmentation of susceptible tissue occurs while con-
currently lavaging and aspirating material from the surgical site. The CUSA
selectively ablates tissues with high water content such as liver paren-
chyma, glandular tissue, and neoplastic tissue. Among CUSA’s benefits, it
provides a very well-defined transection plane, which is useful in situations
of close proximity between tumors and major vascular structures. Also, it
can be used in the cirrhotic as well as noncirrhotic livers and is associated
with a low blood loss and low risk of bile leak.
Noninferiority studies comparing CUSA to the clamp–crush technique
have shown similar outcomes. One of the first randomized studies compar-
ing the ultrasonic dissector versus the clamp–crush technique showed that
the ultrasonic dissector is more frequently associated with tumor expo-
sure at the resection margin and with incomplete appearance of landmark
hepatic veins on the cut surface. Subsequent retrospective studies showed
superiority with respect to blood loss but with longer operating times com-
pared to conventional surgery. Other studies comparing CUSA with water
dissectors (Hydrojet) showed CUSA to have longer transection times but
0002086380.INDD 163 7/15/2014 12:36:02 PM

less blood loss. Several RCTs looking at the effectiveness have not shown
any improvement in operative times, blood loss, or complication rates with
CUSA. Proponents of CUSA highlight the selective vascular isolation and
visualization of the CUSA, the clean operative field, and the fact that it has
been used clinically for close to two decades. Opponents argue that is takes
longer, has similar transection outcome, and is expensive. The incidence
of air embolism with ultrasonic dissectors is also a common phenomenon,
which is usually asymptomatic but may lead to adverse outcomes.
Water-Jet Dissector (ERBEJET 2, ERBE USA Inc., Marietta, GA)

This technique employs a high-pressure water jet to break apart the liver
tissue and selectively isolate small vascular and biliary structures, poten-
tially decreasing blood loss. These vessels and ducts must then be ligated
and divided individually according to preference. Its advantage is hav-
ing a precise delineation of the transection plane, which can be used to
expose major vessels effectively, especially in the context of closely adjacent
tumors. Water-jet dissection does not use thermal energy and thus spares
the surrounding tissue from any thermal damage. Water-jet dissection was
noted in one randomized study to be superior to CUSA with regard to tran-
section speed and blood loss, Pringle time, and operative duration. A ran-
domized trial comparing four techniques of clamp–crush (under pedicle
control), CUSA, Hydrojet, and dissecting sealer was recently conducted.
The clamp–crush method was noted to be superior in transection speed,
blood loss, transfusion rate, number of ligatures, and transection time. The
ultrasonic and water-jet dissection devices have remained a useful tool for
precoagulation dissection and thus do require an additional coagulation
device and cosurgeon of adequate experience to assist during the transec-
tion steps. This requirement can limit efficiency, but overall remains a very
safe and effect method of dissection.
ULTRASONIC CUTTING AND COAGULATING DEVICE

The ultrasonic cutting and coagulating surgical device (Harmonic scalpel,
not to be confused with ultrasonic dissector such as CUSA) converts sound
energy into mechanical energy at the functional end of the instrument.
A piezoelectric crystal in the hand piece generates vibration at the tip of
the active blade at 55,500 times per second. This leads to production of heat
from rupture of hydrogen bonds, which leads to denaturation of proteins
at tissue temperatures of 60°C to 80°C, resulting in coagulum formation
without the desiccation, charring, or smoke caused by conventional elec-
trosurgical methods. The main disadvantages of Harmonic scalpel are the
limited ability to coagulate vessels larger than 3 to 5 mm, increased cost of
disposable instruments, and potential for extensive thermal spread at high
energy levels (level 5) for more than 5 seconds. The quality of vessel seal also
depends on the amount of tissue tension delivered and the setting used,
and in this aspect, it is surgeon dependent.
Use of Harmonic scalpel has become more widespread in the era of
laparoscopic surgery, with varied uses from thyroid surgery to upper gas-
trointestinal (GI) tract procedures. Harmonic scalpel is capable of sealing
vessels from 3 to 5 mm and has shown to have less lateral thermal spread
than conventional electrosurgery and minimal smoke formation or char-
ring. It is also a versatile instrument, allowing the surgeon to dissect, cut,
and coagulate using one instrument. The presence of an active exposed
blade can be both an advantage in mobilizing coapted tissue but can also
be dangerous if unintentional tissue contact with the active blade occurs.
Initial reports used Harmonic scalpel in both open and laparoscopic liver
002086380.INDD 164 7/15/2014 12:36:02 PM

resections, with no biliary leakage reported in a study of 41 patients. In

contrast, a nonrandomized study showed an increased rate of postopera-
tive bile leakage, raising some concerns of its safety sealing biliary struc-
tures. Another large published series comparing clamp–crushing with
ultrasonic plus Harmonic scalpel dissection showed longer operative time,
but with a reduced blood loss and a lower rate of biliary fistula. However,
the retrospective method of the study and the relatively long period of
inclusion may have biased these results against the clamp–crush tech-
nique. The merits of Harmonic device in cirrhotic livers, or in preventing
bile leaks, remain to be proven. Similarly, the device has not been opti-
mized for prolonged 20- to 30-second continuous use during transection,
and thus, care should be taken to avoid collateral heat transfer and to
ensure device integrity.
COMPARISON
The techniques of finger fracture technique or the use of Kelly-clysis along
with individual ligation with clips or sutures for vascular and biliary struc-
tures have long been employed in liver resection. There has since been an
increase in the use of precoagulation, or making the transection plane
avascular with coagulation devices before transection. Examples of preco-
agulation devices are microwave and radiofrequency energy devices, which
ablate and coagulate along the transection plane leaving a rim of avascu-
lar tissue. Broadly speaking, Ligasure, Enseal, and Harmonic devices use
the precoagulation principles to effect subsequent parenchymal division.
This is in contrast to resistance-modulated devices, which fragment low-
resistance tissue (hepatic parenchyma) preserving fibrous (high-resistance)
components such as vessels and bile ducts for subsequent ligation by the
surgeon. These vascular-sparing structures include the ultrasonic dissector
(CUSA) and the Hydrojet devices.
The advent of laparoscopy and its increased use in hepatic surgery led
to an increased need for improved transection technology. The use of inflow
control in laparoscopic surgery is less common than with conventional
techniques due to technical difficulty. Newer transection devices have
allowed parenchymal transection without the need for prolonged inflow
control in laparoscopic procedures.
Although the Kelly-clysis or clamp–crush technique has long been
the cornerstone of parenchymal transection in combination with inflow
control, there is a greater risk of blood loss associated with this technique,
when performed incorrectly. Despite a plethora of published literature
with respect to parenchymal transection devices, no clear consensus of the
optimal device exists. One of the major problems encountered in perform-
ing trials in this setting is rapid advance of technology. Newer variations
of existing technology and refinements in current devices do not present
enough time for planning and conduct of appropriately powered trials.
Again, when these trials do happen, they are quickly outmoded by a more
novel technology. This is due to the constant feedback from the operator
of the devices and manufacturers trying to retain a niche market. Also,
the operative mortality is about 1% in recent studies with a similar low
morbidity of 8%. Therefore, sample size needed for adequately powered
randomized controlled trials using surgery-related mortality or specific
complications as a primary end point is very high, and most single-center
studies are unable to reach those numbers. Multicenter trials are marred
by nonuniformity of technique, which can lead to aberrant interpretations.
At this current point in time, no specific claims of superiority can be made
based on the evidence-based literature of one energy device over another.
0002086380.INDD 165 7/15/2014 12:36:02 PM

Published studies have mixed results and design flaws and are underpow-
ered. In an attempt to resolve this issue, a few meta-analyses have been
performed; these have showed no clear benefit of one device over another
or in fact even over the conventional clamp–crush technique. In this void
of definitive literature or consensus, the individual surgeon’s technique,
prior training, and familiarity with the device are the key to maintain the
current quality standards for hepatic transection and postoperative mor-
bidity and mortality.
In summation, most parenchymal transection devices have certain
advantages and niche users. Ligasure has the most published literature, is
capable of sealing vessels up to 7 mm, has low lateral thermal imprint and
spread, and is easy to use. Similarly, the newer Enseal device has a similar
vessel-sealing ability, with lower thermal imprint, uses compression forces
in addition, and can cut and coagulate at the same time but needs a trained
measured input from the user. Harmonic scalpel is often used in combi-
nation with CUSA, Hydrojet, or the clamp–crush technique, but is not
suitable for controlling larger biliary or vascular structures. Using of resis-
tance-modulated devices, such as CUSA or Hydrojet, provides clean lines of
transection. Proponents argue that in performing hepatectomies in which
small margins are anticipated, the Harmonic scalpel could be beneficial
in achieving R0 resections. Primary precoagulation technology of RF-PT
(Habib) is popular among many centers, with reported bloodless transec-
tion field especially for nonanatomical resections but leaves a large area of
necrosis in its wake. Microwave devices are newer energy platforms with a
smaller zone of necrosis. Some large centers use salient dissecting sealer
often in conjugation with another hemostatic device and prefer the clean
transection with minimal thermal spread. It would be remiss to discuss
parenchymal transection without mentioning hepatectomy with staplers.
Stapler-directed parenchymal transection is safe and fast and leads to less
blood loss and shorter hospital stay. Initial studies have yet to be confirmed,
and the increase in cost may deter some surgeons. The CRUNSH study is
a randomized controlled trial to evaluate stapler hepatectomy versus the
clamp–crush technique with the primary end point being intraoperative
blood loss; this trial has recently been completed.
Most of these devices when used properly are safe and can facilitate
parenchymal transections. Therefore, the choice of hemostatic assist tech-
nology in liver transection should be based on individual surgeon prefer-
ence, training, and practice setting. Surgeons should use mentors, industry
workshops, animal labs, and current literature to train themselves to the
device(s) they use. As a word of caution, some of these powerful energy
devices, if used inappropriately, can have severe consequences. For exam-
ple, in aberrant anatomy, or anatomy distorted by tumors, large vascular or
biliary structures could be easily sealed off with disastrous consequences
(Table 13.1).
Our Perspective
The published literature in hepatic parenchymal transection is heteroge-
neous. Some technologies have been shown to be superior in histologic
models and animal studies, but randomized controlled trials or similar
high-quality clinical data do not exist. At this point, no single energy source
has shown to be superior to others, and it behooves the operator to be aware
of the risks and benefits of the technology that they are using. Currently,
many surgeons are using an “energy device”–clysis technique in which they
replace the Kelly clamp with hemostatic device as the crush–clamp device.
This was originally used with harmonic focus and was described as focus-
clysis but is being used with other devices.
002086380.INDD 166 7/15/2014 12:36:02 PM

0002086380.INDD 167
TABLE
Vessel-Sealing Devices’ Comparison
13.1 Ligasure Enseal PK-Gyrus Harmonic
Mechanism Bipolar Bipolar Bipolar Ultrasonic shear
Vessel size Up to 7 Up to 7 5 5
Lateral thermal spread 2–3 1 2–6 1
Burst pressure (2–3 mm) 744 1,025 910 789
Burst pressure (4–5 mm) 1,261 927 647 390
Burst pressure (6–7 mm) 645 720 284 532
Seal times (2–3 mm) 2.8 4 2 4
Seal times (4–5 mm) 3 6.8 4 3.9
Seal times (6–7 mm) 3.5 8.25 4.5 5.2
Failure rates 0–13.3 0–5.8 11–40 0–22
Smoke/vapor (ppm) 12.5 ± 3.6 21.6 ± 5 74.1 ± 12 2.9 ± 0.6
Operator-dependent sealing No Yes No Yes
Tactile feedback during actuation No Yes No No
Learning curve Minimal Moderate Minimal Moderate
7/15/2014 12:36:02 PM
Our current practice is to use a similar technique in open and laparo-

scopic as well as in cirrhotic livers. The mobilization of the liver is done with
combined use of sharp Bovie dissection and Enseal. Pedicle clamping is the
exception rather than the norm, and this helps reduce blood loss in patients
who have steatohepatitis or cirrhosis. After performing intraoperative
ultrasound and marking the planned line of liver transection, a combined
crush and vessel sealing is performed with the Enseal device. Dissection is
advanced from the liver surface in 2- to 3-mm layers and 3 cm depth, avoid-
ing making “tunnels” or “holes.” The inflow and outflow tract is controlled
with a single vascular stapler. Our preference for the Enseal device is due
to its high-burst pressures, low lateral thermal spread, ability to engage the
coagulation without locking or clamping down the instrument, uniform
application of pressure along the length of the jaws, comparative cost effi-
cacy and the ability to use the same device across a spectrum of liver resec-
tions including open or laparoscopic cases.
Suggested Readings
Gurusamy KS, et al. Techniques for liver parenchymal transection in liver resection.
Cochrane Database Syst Rev 2009;(1):CD006880.
Kluger MD, Cherqui D. Minimally invasive techniques in hepatic resection. In:
Jarnagin WR, Blumgart WH (eds.), Blumgart’s surgery of the liver, pancreas and
biliary tract, 5th ed. Philadelphia, PA: Elsevier, 2012.
Newcomb WL, et al. Comparison of blood vessel sealing among new electrosurgical
and ultrasonic devices. Surg Endosc 2009;23(1):90–96.
Pamecha V, et al. Techniques for liver parenchymal transection: a meta-analysis of
randomized controlled trials. HPB(Oxford) 2009;11(4):275–281.
Rahbari NN, et al. Clamp-crushing versus stapler hepatectomy for transection of the
parenchyma in elective hepatic resection (CRUNSH)—a randomized controlled
trial (NCT01049607). BMC Surg 2011;11:22.
002086380.INDD 168 7/15/2014 12:36:02 PM

14 Hepatocellular Carcinoma
Mashaal Dhir and Chandrakanth Are
INTRODUCTION
Worldwide, hepatocellular carcinoma (HCC) is the fifth most common can-
cer and a leading cause of cancer-related mortality. The major risk factors
for HCC include presence of cirrhosis and chronic viral hepatitis due to hep-
atitis C virus (HCV) and hepatitis B virus (HBV). HBV is the predominant
risk factor in the developing world, whereas HCV is the major risk factor in
the United States. HBV and HCV coinfection as well as human immunodefi-
ciency virus and HCV coinfection may accelerate the development of HCC.
Nonalcoholic fatty liver disease or nonalcoholic steatohepatitis (NASH) is
commonly seen in obese and diabetic individuals and is currently on the
rise worldwide and may become a major cause of HCC in future. HCV and
NASH account for majority of cases of HCC in the United States. Alcohol,
although not a mutagen itself, can cause HCC through alcohol-induced
cirrhosis and can act synergistically with other risk factors, including viral
hepatitis. Aflatoxin B1 exposure, autoimmune disorders such as primary
biliary cirrhosis and autoimmune hepatitis, and several inherited liver dis-
orders such as hemochromatosis, Wilson disease, alpha1-antitrypsin defi-
ciency, hereditary tyrosinemia, and glycogen storage disorders, although
rare, can also predispose to HCC.
HCC is the result of chronic injury to the liver cells leading to repeated
cycles of cell damage and proliferation with accruals of several genetic and
epigenetic changes. Major molecular alterations in HCC include TP53 muta-
tion (approximately 50%), Wnt–beta-catenin signaling pathway mutations
(20% to 40%), activation of the insulin-like growth factor signaling pathway,
and PI3/PTEN/AKT signaling pathway activation. The PI3/PTEN/AKT
pathway plays a role in cell invasion through activation of MMP-9; there-
fore, kinases in this pathway could prove to be exciting therapeutic targets.
SURVEILLANCE AND DIAGNOSIS

Several societies including the American Association for the Study of Liver
Diseases (AASLD) and the National Comprehensive Cancer Network have
published guidelines for screening and surveillance of HCC. The AASLD
prefers the use of term surveillance over screening for early detection of
HCC as it involves repeated use of screening test at repeated intervals.
Surveillance for HCC is indicated in all cirrhotic patients. Surveillance is
also recommended for other specific groups such as hepatitis B carriers
with positive family history of HCC and those with high HBV DNA levels.
Both radiologic and serologic tests are used for surveillance of HCC in
the above-mentioned populations. Ultrasound (US) examinations and alfa-
fetoprotein (AFP) levels are the most widely used methods for surveillance.
A surveillance interval of 6 months is based on tumor doubling times of
approximately 6 months.
169
0002086381.INDD 169 7/14/2014 7:05:55 PM

Among the serologic methods, AFP is the most commonly used diag-
nostic test although sensitivity and specificity vary with the size of the
nodule and the cutoff values utilized. Sensitivity and specificity range from
55% to 60% and 88% to 90% for a cutoff value of 20 μg/L compared to 17%
and 99.4%, respectively, for a cutoff value of 400 μg/L. In addition, AFP may
be elevated in patients with intrahepatic cholangiocarcinoma. Therefore,
per AASLD, diagnosis should rest on the radiologic and histopathologic
criteria, and AFP alone is considered as an inadequate screening test for
HCC. US is the most commonly recommended test overall for surveillance
with sensitivity (65% to 80%) and specificity (48% to 94%) varying depend-
ing on the size of the nodule. Importantly, US is operator dependent, and
all focal lesions on US should be verified using four-phase multidetector
CT (MDCT) or a dynamic contrast-enhanced MRI as these modalities have
higher sensitivity (89%) and specificity (99%) for lesions greater than 1 cm.
A contrast-enhanced study is required to demonstrate the typical “wash-
out” of the contrast seen in HCC. HCC has arterial blood supply exclusively,
whereas the rest of the liver has both portal venous and arterial supply. In
the arterial phase, HCC enhances more than the liver as it has only arterial
supply, whereas arterial contrast in nontumorous liver is diluted by the por-
tal venous blood. In the venous phase, HCC enhances less than the liver as
it has no venous blood supply and the arterial blood flowing through HCC
no longer contains contrast, whereas the nontumorous liver is enhanced by
contrast in the venous blood. This phenomenon is known as “washout” and
is characteristic of HCC. A four-phase study (unenhanced, arterial, venous,
and delayed—MDCT or dynamic MRI) is required to document this radio-
logic finding. Figure 14.1 summarizes the AASLD-recommended diagnostic
algorithm for suspected HCC (Fig. 14.1).
HCC
< 1 cm > 1 cm
Repeat US at 3 months 4-phase MDCT/ dynamic

contrast enhanced MRI
Arterial hypervascularity AND

venous or delayed phase
washout
Growing/changing
character Stable
Other contrast enhanced

Yes study (CT or MRI) No
Investigate Arterial hypervascularity AND

according to size HCC venous or delayed phase Biopsy
washout
Yes No
FIGURE 14.1 Algorithm for investigation of small nodules found on screening in patients at
risk for HCC. Adapted from AASLD practice guidelines 2010. Bruix et al. 2010, with permission.
002086381.INDD 170 7/14/2014 7:05:56 PM

Chapter 14 / Hepatocellular Carcinoma 171
STAGING SYSTEMS FOR HCC

HCC arises in the background of a preneoplastic cirrhotic liver, and both
HCC and the underlying cirrhosis along with the patient’s condition may
affect the mortality of the patient. Similarly, the biology of the disease and
the risk factors are different in the East and West. HCC is the one of the few
tumors that can be treated by both resection and transplantation. Because
of these unique challenges, it is difficult to devise a uniform staging system
and, consequently, several staging systems for HCC exist. An ideal staging
system should take into account the tumor burden, functional state of the
liver, and performance status of the patient. The Barcelona Clinic Liver
Cancer (BCLC) staging system is currently the most widely utilized staging
system worldwide as it includes tumor stage, liver functional status, and
physical status of the patients and links the various stages to treatment
modalities and estimated life expectancies. Figure 14.2 summarizes the
BCLC staging system (Fig. 14.2).
TREATMENT OF HCC
There are several treatment options available to patients diagnosed with
HCC. The treatment algorithms are guided by the extent of tumor burden
(extrahepatic spread, vascular invasion, number of tumors, tumor diam-
eter), severity of underlying liver dysfunction, size of future liver remnant
(FLR), functional status, and comorbidities of the patient.
Transplantation and resection continue to be the major curative-
intent therapeutic options available to patients with HCC. Patients with
early-stage disease (i.e., HCC falling within the Milan criteria—solitary
lesion ≤ 5 cm or ≤ 3 lesions with the largest diameter ≤ 3 cm and absence of
macroscopic vascular invasion or extrahepatic disease) and advanced cir-
rhosis including Child-Pugh class B/C and portal hypertension are thought
to be candidates for transplantation, whereas resection remains the treat-
ment of choice for patients without any significant underlying liver disease.
Ablative treatment, chemoembolization, and systemic therapies play an
important role as well.
HCC
Stage 0 Stage A–C Stage D

PS 0, Child-Pugh A PS 0–2, Child-Pugh A-B PS >2, Child-Pugh C
Very early stage (0) Early stage (A) Intermediate stage (B) Advanced stage (C) Terminal
Single < 2 cm Single or 3 nodules < 3 cm, PS 0 Multinodular, PS 0 Portal invasion, N1, M1, PS 1–2 stage (D)
Single 3 nodules ≤3 cm
Portal pressure/ bilirubin

Increased Associated diseases
Normal No Yes
Resection Liver Transplantation RFA TACE Sorafenib

Symptomatic
treatment
Curative treatments Palliative treatments
FIGURE 14.2 Barcelona Clinic Liver Cancer (BCLC) staging system. PS, performance sta-
tus; N, nodal status; M, metastases. Adapted from AASLD practice guidelines 2010. Bruix
et al. 2010, with permission.
0002086381.INDD 171 7/14/2014 7:05:57 PM

Liver Resection
Liver resection may be considered the primary treatment modality in
some patients with HCC in the absence of cirrhosis. Patients with well-
compensated cirrhosis, that is, Child-Pugh A cirrhosis, absence of portal
hypertension, and Model for End-Stage Liver Disease (MELD) score of less
than 10, may be considered for resection as well. Resection is usually con-
traindicated in patients with advanced liver disease, that is, Child-Pugh C
and majority of patients with Child-Pugh B with portal hypertension.
Liver resection provides specimen for pathologic confirmation of
diagnosis and can also help to ascertain other pathologic tumor factors,
which may help provide information on the biology of tumor. Unlike ortho-
topic liver transplant (OLT), liver resection is not usually limited by tumor
size/number or macrovascular invasion, and there is no waiting time.
Unfortunately, the recurrence rates tend to be higher as resection does not
address the premalignant potential of the FLR.
Preoperative Assessment of Liver Function

Careful preoperative assessment of liver function is important since post-
operative liver failure in patients undergoing liver resection is associated
with high mortality and morbidity. A complete metabolic panel is one of
the most basic tests that can be performed to assess the liver function.
Candidates who have evidence of active hepatitis as indicated by elevated
bilirubin, AST, or ALT may be poor candidates for resection. In the Western
world, the Child-Pugh scoring system has been used traditionally to esti-
mate the hepatic functional reserve prior to liver resection. Patients with
Child-Pugh class A and those with highly select Child-Pugh B are consid-
ered candidates for resection. More recently, the MELD score has been
applied to preoperative selection of HCC resection candidates, as well.
Portal hypertension is considered to be present when hepatic venous
pressure gradient is greater than 10 mm Hg. Portal hypertension increases
the risk of major postoperative complications in the form of variceal bleed-
ing, endotoxemia, and postoperative hepatic decompensation. Therefore,
clinical or radiologic criteria to diagnose portal hypertension are a key step
in the evaluation of candidates for resection. Platelet count of less than
100,000/μL associated with splenomegaly, ascites requiring drug treat-
ment or intervention, esophagogastric varices on upper endoscopy, and
presence of abdominal wall collaterals may point toward significant portal
hypertension.
In the East, further functional assessment of the liver is performed
preoperatively most commonly using indocyanine green (ICG) clearance.
Functionally, ICG is taken up by hepatocytes and excreted in the bile in an
adenosine triphosphate-dependent fashion. Therefore, its clearance from
the systemic circulation is an indicator of hepatic function. The amount of
ICG remaining in the bloodstream of a patient with normal liver function 15
minutes after the injection should be less than 10%. The ICG clearance test
is not commonly used in the United States.
Evaluation of Future Liver Remnant Volume

Major or extended hepatectomy may lead to an inadequate FLR that can
be associated with significant risk of hepatic insufficiency and subse-
quent mortality and morbidity. Although the risk of hepatic insufficiency
is determined by several factors as highlighted above, the size of FLR con-
tinues to be one of the major determinants of postoperative hepatic failure.
Precise measurements of hepatic volume are needed before operating on
any patient that is likely to be left with an inadequate FLR. The FLR can be
calculated using three-dimensional (3D) CT volumetry. FLR has been used
002086381.INDD 172 7/14/2014 7:05:57 PM

as a surrogate to predict postoperative outcomes. In patients with normal

liver function, FLR of at least 20% is recommended. For patients with cir-
rhosis and those treated with systemic chemotherapy, a higher FLR is rec-
ommended (40% for cirrhosis, 30% after systemic chemotherapy) due to the
underlying liver dysfunction.
Portal Vein Embolization

A small FLR may increase the risk of posthepatectomy liver failure (PHLF).
However, this can be avoided by inducing ipsilateral atrophy of the tumor-
bearing liver and compensatory hypertrophy of the FLR by selectively
occluding the blood flow to the tumor-bearing part of the liver. Hypertrophy
of the nonembolized, non–tumor-bearing liver could be secondary to
increased portal blood flow through redistribution as well as release of
cytokines such as hepatocyte growth factor and transforming growth fac-
tor α and β. In general, portal vein embolization (PVE) is offered to patients
with (1) normal liver function and an FLR of 25% to 30% and (2) with com-
promised liver function such as postchemotherapy liver damage, cirrhosis/
fibrosis, cholestasis, and FLR of 35% to 40%. In the post-PVE period, the
FLR undergoes rapid hypertrophy in the ensuing 3 to 4 weeks. In patients
who have diabetes or cirrhosis, the hypertrophy may be delayed, and an
additional 3 to 4 weeks may be required to assess the complete response.
A small percentage of patients undergoing PVE may develop severe com-
plications in the form of severe cholangitis, large abscesses, sepsis, portal
venous, or mesentericoportal venous thrombosis, precluding liver resec-
tion. In a different strategy, PVE embolization can be utilized as a “stress
test” for the liver. Patients who undergo sufficient hypertrophy may do well
with resection, whereas those with insufficient hypertrophy are more likely
to experience complications and posthepatectomy liver failure. Similarly,
any disease progression seen during the period of PVE may indicate aggres-
sive tumor biology, and resection may not change the course of the disease
in these patients.
Outcomes
Most major centers report a perioperative mortality of less than 5% after
liver resection for HCC. Cirrhosis and portal hypertension may impair liver
regeneration, increase intraoperative blood loss requiring transfusion, and
increase the risk of liver failure. PHLF is one of the most serious complica-
tions after liver resection and is reported in approximately 5% to 10% of
patients. Careful patient selection, limiting resection to patients with well-
compensated cirrhosis (Child A) in the absence of portal hypertension, and
parenchymal-sparing resections are all strategies to help prevent postop-
erative liver failure. Additionally, intraoperative blood loss can be limited
by reducing the central venous pressure and intermittent portal clamping.
The overall 5-year survival after hepatic resection ranges from 25% to
50% depending on the size and number of HCC nodules, vascular invasion,
and level of AFP. Multifocal HCC and major vascular invasion are associ-
ated with poor survival. Resection does not address the residual (usually
cirrhotic) liver whose function may continue to worsen and in which new
tumors may develop. In patients with small solitary HCC and well-preserved
liver function, the 5-year survival may exceed 50%.
Orthotopic Liver Transplant

Generally speaking, OLT is the preferred treatment approach for cirrhotic
patients falling within the Milan criteria as long-term outcomes are compa-
rable to patients with similar stage of cirrhosis in the absence of malignancy.
The Milan criteria include (1) single HCC less than 5 cm in size, (2) three or
0002086381.INDD 173 7/14/2014 7:05:57 PM

fewer nodules each less than 3 cm in size, and (3) absence of extrahepatic
spread or major vascular invasion. Extrahepatic staging should include CT
of the chest and CT/MRI of the abdomen and pelvis. In patients with cir-
rhosis undergoing transplantation for early stage HCC, the 5- and 10-year
survival may vary from 60% to 80% and 50% to 60%, respectively. However,
most studies do not take into account the mortality while awaiting liver
transplant or dropout from the waitlist as some patients experience disease
progression.
At most centers, organ allocation is based on the MELD score.
Researchers at the University of San Francisco have proposed expanded
size criteria for HCC in transplant candidates (UCSF criteria), that is, single
HCC less than 6.5 cm, maximum of three total tumors with none greater
than 4.5 cm, and cumulative tumor size less than 8 cm. The overall 1- and
5-year survival rates were 90% and 75%, respectively, for patients meeting
the UCSF criteria. Therefore, selected patients with stage 3 disease may be
candidates for transplantation.
Downstaging
Many times, HCC is detected at an advanced stage, and such patients are
no longer candidates for OLT. The goal of downstaging is to decrease the
size and number of tumors in patients who do not meet the criteria for
transplantation on initial evaluation. Liver transplantation after successful
downstaging should achieve a 5-year survival comparable to HCC patients
undergoing transplantation without downstaging.
No clearly defined upper limits for size and number of lesions qualify
or preclude a patient downstaging. However, the presence of extrahepatic
disease and major vascular invasion are often considered as contraindica-
tions for downstaging. Transarterial chemoembolization (TACE) and radio-
frequency ablation (RFA) are commonly used downstaging techniques.
In patients who are successfully downstaged, a minimum wait period of
3 months is recommended prior to transplantation. Failure of downstag-
ing may be defined as (1) before listing— failure to achieve listing criteria,
tumor progression with or without development of major vascular inva-
sion, and extrahepatic spread or tumor size or spread beyond the inclusion
criteria—and (2) after listing—tumor progression requiring delisting.
Management of Patients on the Waiting List

Dropouts of HCC patients on the waiting list are common due to cancer
progression or other medical reasons. Current consensus guidelines recom-
mend at least three monthly monitoring of listed patients using imaging
(dynamic CT, dynamic MRI, or contrast-enhanced US) and AFP measure-
ments to identify those who undergo disease progression. Bridging strat-
egies with locoregional therapies may be used to decrease tumor-related
dropout rates and are usually beneficial for patients with a wait time of 6
months or more. The most common bridging strategies include TACE, RFA,
a combination of TACE and RFA, 90yttrium radioembolization, or hepatic
resection.
LOCOREGIONAL THERAPIES
Locoregional therapies in the treatment of HCC are aimed at inducing
selective tumor necrosis. These therapies can be broadly divided into two
categories, that is, ablation versus embolization. The extent of tumor necro-
sis induced by locoregional therapies can be assessed by imaging as well as
biochemical criteria. Posttreatment dynamic CT/MRI and AFP can be used
to assess the response to locoregional therapies.
002086381.INDD 174 7/14/2014 7:05:57 PM

Ablation
Liver cancer cells can be destroyed by chemical substances such as ethanol
or acetic acid or by modulation of temperature of the cancer cells by radio-
frequency, microwaves (MW), laser, or cryoablation.
Percutaneous ethanol injection (PEI) and RFA are the two most com-
monly used ablative treatments. Any ablative treatment can be performed
by laparoscopic, percutaneous, or open technique. In patients who receive
PEI, distribution of ethanol may be blocked by intratumoral septa and/or
tumor capsule, resulting in heterogenous distribution. As a result of this
anatomic compartmentalization, curative capacity of PEI decreases as
the size of tumor increases. PEI can achieve complete tumor necrosis in
almost all patients with tumors smaller than 2 cm and approximately 70%
of patients with tumors smaller than 3 cm.
In contrast, RFA causes destruction of the tumors and surrounding
liver parenchyma through high-frequency alternating electrical current,
generating temperatures in the range of 70°C to 105°C, which leads to coag-
ulative necrosis. Ablation also generates a margin of ablated nontumoral
tissue, which might eliminate any undetected satellite nodules. One of the
major limitations of RFA includes inability to ablate lesions located close to
major vascular structures, which act as a heat sink.
In patients with small HCC (< 3 cm) and well-preserved liver func-
tions, RFA may be considered the first-line ablative treatment. In contrast,
PEI may be reserved for lesions not suitable for RFA such as pericholecys-
tic lesions or those near the hepatic hilum. The efficacy of percutaneous
ablative treatments may be assessed by contrast-enhanced CT or MRI at
1 month after therapy. Absence of contrast uptake at 1 month is associated
with tumor necrosis, and persistence of contrast uptake indicates treat-
ment failure.
Proximity to large blood vessels can lead to absorption of heat by the
rapidly flowing blood (heat-sink phenomenon); this heat sink can lead to
reduced efficacy and is a relative contraindication to RFA. There is higher
chance of tumor rupture after RFA of lesions located near the liver capsule.
Caution is advised for lesions located near the major bile ducts, bowel,
gallbladder, diaphragm, heart, and stomach as these structures may be
damaged.
Other Ablative Strategies

MW are electromagnetic waves that can produce thermal effects leading
to tumor destruction. MW ablation has some advantages over RFA, such
as a higher temperature at the target tissue, a shorter duration of therapy, a
less pronounced cooling effect of adjacent vessels (less heat-sink effect), an
effective treatment of cystic lesions, and lack of need for neutral electrodes,
thus avoiding the risk of related skin burning. Irreversible electroporation
(IRE) is a recent addition to the armamentarium of ablative strategies. In
IRE, small electrical pulses cause disintegration of the cell membranes lead-
ing to cell death without producing the thermal effect. Functionally, IRE
causes either definite cell death or no death, so the area of damage is very
well defined. In theory, and unlike RFA or MW, there is little damage to the
surrounding vessels, bile ducts, or bronchioles.
Embolization
Embolization is based on catheter-based infusion of particles targeted at the
arterial branch of the hepatic artery feeding the portion of the liver in which
the tumor is located. The liver has dual blood supply derived from the portal
vein and hepatic artery. Tumors such as HCC are supplied p redominantly
by the hepatic artery and are hypervascular resulting from increased blood
0002086381.INDD 175 7/14/2014 7:05:57 PM

flow to the tumor compared to normal liver tissue. The blood supply to the
tumor is carefully isolated, and embolization is performed to the subseg-
ment, segment, or lobe of the liver containing the tumor. Nontarget embo-
lization to the liver may result in serious liver damage. Embolization can
be performed using bland particles, chemotherapy or drug-eluting beads
(DEB-TACE), or beads tagged with a radionuclide (90Yttrium).
Embolization plays multiple roles in the management of HCC. Major
indications include (1) palliative treatment for patients with unresectable/
inoperable disease with tumors not amenable to ablation treatment only
and minimal or absence of large-volume extrahepatic disease; (2) treat-
ment for postresection intrahepatic recurrence; (3) primary treatment for
ruptured HCC; (4) adjuvant treatment to prevent postoperative recurrence;
(5) neoadjuvant treatment for large resectable HCC to reduce tumor vol-
ume; (6) bridging treatment to inhibit tumor growth in patients awaiting
transplant.
Selection Criteria
Preoperative procedure workup includes evaluation of hepatic functional
reserve and baseline tumor markers and cross-sectional imaging to assess
the size, number, and macroscopic vascular invasion of the hepatic or por-
tal vein. Assessment of the patient’s comorbidities, functional status, and
evaluation for metastatic disease also play important roles in treatment
planning. The endpoint of treatment includes stasis in tumor-feeding arter-
ies and appearance of embolization material in the peritumoral portal vein
tributaries. Patients are routinely followed at 2- to 4-week intervals with
liver function tests and tumor marker assays. Similar to ablative treatment,
tumor necrosis by embolization can be assessed by contrast uptake on
dynamic CT/MRI 4 to 8 weeks posttreatment.
OTHER STRATEGIES
External beam radiation therapy (EBRT) allows focal administration of
high-dose radiation to HCC tumors in patients with unresectable tumors
or inoperable liver disease. Unlike ablation treatment, EBRT is not limited
by tumor location.
Systemic therapies including sorafenib are being explored. Sorafenib
is an oral multikinase inhibitor that suppresses tumor cell proliferation
and angiogenesis and has been evaluated for the treatment of advanced or
metastatic HCC.
In conclusion, treatment of HCC is evolving. Resection and transplan-
tation remain major surgical options. Locoregional therapies including
ablative treatments and transarterial embolization are increasingly utilized
in the management of HCC patients. The role of systemic therapies is evolv-
ing. The evolution of these varied treatment strategies combined with a bet-
ter understanding of the biology of the disease should contribute to further
improvements in outcomes for patients with HCC.
Suggested Readings
Belghiti J, Carr BI, Greig PD, et al. Treatment before liver transplantation for HCC. Ann
Surg Oncol 2008;15(4):993–1000.
Benson AB, III, Abrams TA, Ben-Josef E, et al. NCCN clinical practice guidelines in
oncology: hepatobiliary cancers. J Natl Compr Canc Netw 2009;7(4):350–391.
Bruix J, Sherman M. Management of hepatocellular carcinoma: an update. Hepatology
2011;53(3):1020–1022.
Clavien PA, Lesurtel M, Bossuyt PM, et al. Recommendations for liver transplanta-
tion for hepatocellular carcinoma: an international consensus conference report.
Lancet Oncol 2012;13(1):e11–22.
002086381.INDD 176 7/14/2014 7:05:57 PM

El-Serag HB. Epidemiology of viral hepatitis and hepatocellular carcinoma.

Gastroenterology 2012;142(6):1264–1273 e1.
Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma. Lancet 2012;379(9822):
1245–1255.
Jarnagin W, Chapman WC, Curley S, et al. Surgical treatment of hepatocellular carci-
noma: expert consensus statement. HPB (Oxford) 2010;12(5):302–310.
Jemal A, Bray F, Center MM, et al. Global cancer statistics. CA Cancer J Clin
2011;61(2):69–90.
Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib in advanced hepatocellular carci-
noma. N Engl J Med 2008;359(4):378–390.
Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation for the treatment
of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med
1996;334(11):693–699.
Rahbari NN, Garden OJ, Padbury R, et al. Posthepatectomy liver failure: a definition
and grading by the International Study Group of Liver Surgery (ISGLS). Surgery
2011;149(5):713–724.
Rahbari NN, Mehrabi A, Mollberg NM, et al. Hepatocellular carcinoma: current man-
agement and perspectives for the future. Ann Surg 2011;253(3):453–469.
Schwarz RE, Abou-Alfa GK, Geschwind JF, et al. Nonoperative therapies for combined
modality treatment of hepatocellular cancer: expert consensus statement. HPB
(Oxford) 2010;12(5):313–320.
0002086381.INDD 177 7/14/2014 7:05:57 PM

15 Management of Patients with
Colorectal Liver Metastases
Danielle A. Bischof and Timothy M. Pawlik
ETIOLOGY
In the United States, approximately 140,000 people are diagnosed with
colorectal cancer (CRC), and 50,000 people die of CRC, yearly. CRC cur-
rently represents the second leading cause of cancer deaths in the United
States. During the course of their disease, half of patients with CRC will
develop colorectal liver metastases (CRLM). The liver is the only site of
metastatic disease in 30% of patients who develop CRLM. These patients
are potential candidates for curative-intent resection of CRLM. Without
treatment, the median survival for patients with CRLM is 6 to 12 months.
With improvements in systemic therapy, the median survival for patients
with unresectable metastatic disease has improved from 12 months to up
to 28 months. Despite this, the survival of patients with CRLM who do not
undergo complete resection remains low at about 5% to 10% at 5 years.
Complete resection remains the only therapy for CRLM with the potential
for long-term cure.
Risk factors for CRC include age, obesity, physical inactivity, alcohol
consumption, diet high in red meat, and smoking. Hereditary and medical
risk factors for CRC include a personal or family history of CRC or colorec-
tal adenomas and a personal history of inflammatory bowel disease and
genetic syndromes such as Lynch syndrome and familial adenomatous
polyposis.
DIAGNOSIS
Diagnosis of CRLM is most common on staging cross-sectional imaging at
the time of initial diagnosis or during surveillance after treatment of the
primary. All patients who are being considered for surgery for CRLM should
undergo thorough preoperative assessment to assess the extent of intrahe-
patic disease as well as to rule out extrahepatic metastases. Patients should
be assessed with computerized tomography (CT) or magnetic resonance
imaging (MRI) of the liver. Preoperative imaging of the liver should char-
acterize the number of liver lesions, their location, and their proximity to
major vascular and biliary structures. In addition, all patients should have
imaging of the chest and a colonoscopy within the last 6 to 12 months to
rule out extrahepatic metastatic disease and local recurrence or a meta-
chronous primary tumor, respectively.
On contrast-enhanced CT scan, liver lesions typically appear hypode
nse and are best seen on the venous phase. Dual-phase helical CT has been
reported to have a sensitivity of 69% to 71% and a specificity of 86% to 91%
for detecting and characterizing CRLM. With newer multiple detector heli-
cal CT, the sensitivity of identifying liver metastases now approaches 80%
to 90% as higher image resolution is achieved with the thin collimation
possible with multidetector row CT. Thin-slice multidetector CT scan also
178
002086382.INDD 178 7/14/2014 4:25:53 PM

Chapter 15 / Management of Patients with Colorectal Liver Metastases 179
provides excellent resolution of vascular and biliary structures to aid opera-

tive planning. Limitations of CT scan are that sensitivity is low for CRLM
less than 1 cm in diameter and for detecting peritoneal-based extrahepatic
metastases. In addition, the sensitivity of CT for CRLM appears to decrease
after patients receive preoperative chemotherapy—especially in the set-
ting of steatosis or steatohepatitis, which can result in the background
liver appearing darker. This results in less contrast between the liver paren-
chyma and the hypovascular metastasis and can hinder detection of CRLM.
CT scan can detect extrahepatic metastases, such as lung metastases.
At the very minimum, all patients with CRLM should undergo a chest x-ray
prior to surgery. The role of chest CT in imaging patients with CRLM remains
controversial among those patients with a normal chest x-ray. Two reports
demonstrated a diagnostic yield of only 4% to 5% of CT chest in patients with
a normal chest x-ray. Chest CT has high sensitivity but is hindered by low
specificity in ruling out pulmonary metastases. Despite this, the NCCN rec-
ommends that a staging CT of the chest be performed in patients with CRC.
Contrast-enhanced MRI with agents such as gadolinium and feru-
moxide has an accuracy of detecting malignant liver tumors of 80% to 90%.
Limitations of MRI are long scanning times and low sensitivity for detect-
ing extrahepatic disease. MRI appears to be superior to CT in the setting
of chemotherapy-induced steatosis or steatohepatitis. A number of fat-
suppressing techniques including chemical shift imaging and fat saturation
techniques can be used to compensate for steatosis in this setting.
Positron emission tomography (PET) and PET/CT have also been
increasingly used to assess patients with CRLM for extrahepatic disease. In
contrast to CT and MRI, PET offers information on the metabolic activity of
the tumor based on the uptake of 18-F fluoro-2-deoxy-d-glucose (18-FDG).
The superiority of PET in staging patients with suspected liver metastases
for extrahepatic disease has been demonstrated in several meta-analyses.
Use of PET has also been associated with decreased rates of nontherapeutic
laparotomy and a change in treatment plan in up to one-third of patients.
PET/CT merges images from PET with CT, which enhances the sensitivity
from 75% to 89% when compared with PET alone. In one study, PET/CT
identified additional lesions in 32% of patients and changed clinical man-
agement in 24%. Several consensus statements have advocated for PET/
CT in the preoperative staging of patients with CRLM. PET/CT is more
effective in identifying metastatic disease prior to starting chemotherapy.
Preoperative chemotherapy impairs glucose uptake in tumor cells due to
diminished hexokinase activity, thereby decreasing the sensitivity of FDG-
PET. Following the initiation of preoperative chemotherapy, a decrease in
avidity on PET/CT has been demonstrated to be an independent predic-
tor of long-term outcome. Complete metabolic response of a metastasis on
FDG-PET after preoperative chemotherapy is, however, an unreliable indi-
cator of complete pathologic response. In general, a baseline PET/CT prior
to chemotherapy administration is recommended in order to accurately
ascertain all metabolically active sites of disease.
In a meta-analysis comparing CT, MRI, and PET for detection of CRC
metastases, PET was the most accurate and sensitive modality for detecting
CRC metastases on a per-patient basis. On a per-lesion basis, CT, MRI, and
PET were comparable in detecting CRC metastases. In contrast, MRI appears
to be the best modality for detecting CRLM in patients who have received
chemotherapy. In a recent meta-analysis, the diagnostic performance of
MRI, CT, and PET/CT was compared in a pooled analysis of patients treated
with preoperative chemotherapy. The pooled sensitivity of MRI for imaging
CRLM after preoperative chemotherapy was 85.7% compared with a sensi-
tivity of only 69.9% for CT, 54.5% for PET, and 51.7% for PET/CT.
0002086382.INDD 179 7/14/2014 4:25:53 PM

A multimodality imaging strategy for preoperative staging of patients

with CRLM, including FDG-PET and contrast-enhanced CT chest and abdo-
men, is probably the most common approach. Contrast-enhanced MRI should
be considered over CT, especially among all patients who have had a signifi-
cant amount of prior chemotherapy and have evidence of steatosis. MRI can
also be useful as an adjunct to CT in patients with equivocal lesions seen on CT.
MANAGEMENT
Improvements in patient selection and perioperative management have
resulted in improvement in the mortality after liver resection from 10% to 20%
before 1980 to approximately 1% currently. Current estimates of mortality,
however, often reflect single-institution experiences from high-volume cen-
ters and may be misleading. In a report looking at mortality after hepatic resec-
tion from the National Inpatient Sample, the adjusted perioperative mortality
for hepatectomy was 5.6%. Patient selection is also critical in ensuring good
postoperative outcomes. Patient comorbidities including coronary artery
disease, renal failure, and heart failure as well as measures of physiologic fit-
ness including the American Society of Anesthesia and the Acute Physiology
and Chronic Health Evaluation scores are predictive of risk of postoperative
complications. Age has not been shown to be an independent risk factor for
increased complications in liver resection, and chronologic age alone should
not be a contraindication to liver resection. The preoperative evaluation
should aim to identify patients who do not have prohibitive operative risk for
hepatectomy and to refer patients who have modifiable risk factors to appro-
priate providers to have these risk factors addressed preoperatively.
A number of clinicopathologic factors that are associated with patient
survival after hepatectomy for CRLM have been identified. These include
stage, nodal status of the primary, disease-free interval from the primary
to development of CRLM, the number and distribution of CRLM, preopera-
tive carcinoembryonic antigen (CEA), and the presence of extrahepatic dis-
ease. A number of these factors were combined to form a clinical prognostic
score by Fong et al. using data derived from 1,001 patients undergoing liver
resection for CRLM. The clinicopathologic factors that comprise the Fong
score include node-positive primary disease, disease-free interval from pri-
mary to metastases less than 12 months, more than one hepatic metastasis,
largest hepatic metastases greater than 5 cm, and CEA level greater than
200 ng/mL. In this scoring system, one point is assigned for each criterion,
and the total score is predictive of patient outcome. Although these factors
are predictive of patient outcome, patients with one or more poor prog-
nostic factors can still derive a substantial survival benefit from hepatic
resection. All patients with resectable CRLM should thus be considered as
surgical candidates and be fully evaluated for resection.
The criteria for resectability of CRLM have changed significantly
over the past 20 years. In the past, the presence of more than three hepatic
metastases, bilobar disease, an anticipated surgical margin of less than
1 cm, hilar lymphadenopathy, or extrahepatic disease was considered either
absolute or relative contraindications to resection for CRLM (Table 15.1).
Each of these criteria has been challenged, and the current definition
of resectability for CRLM is determined by the surgeon’s ability to resect
all sites of disease while leaving an adequate future liver remnant (FLR).
Specifically, resectability is defined by four criteria:
1. An R0 resection of both the primary lesion and any extrahepatic disease
sites must be technically feasible.
2. At least two adjacent liver segments need to be spared.
002086382.INDD 180 7/14/2014 4:25:53 PM

TABLE
Changing Definition of Resectability for Colorectal Liver
15.1 Metastases
Old New
Number of metastases < 4 Any number, provided an R0
resection can be achieved
Distribution of Unilobar Two contiguous segments of
metastases liver must be preserved
Size of metastases < 5 cm Any size, provided an
adequate FLR is preserved
Presence of None Allowed, provided all disease
extrahepatic disease can be resected with an
R0 resection
3. Vascular inflow and outflow and biliary outflow to the remaining liver
segments must be spared.
4. The FLR must be of adequate volume to account for any underlying
hepatic dysfunction.
With expanded criteria of resectability for CRLM, a number of strate-
gies for improving resectability have emerged, with the goal of increasing
the number of patients eligible for hepatic resection for CRLM. These strat-
egies fall into three broad categories: increasing hepatic reserve, decreasing
tumor size, and using combined modality local therapy.
Portal vein embolization (PVE) has been used to increase hepatic
reserve in patients undergoing hepatectomy for CRLM (Fig. 15.1). In
patients with normal hepatic function, an FLR of 20% to 30% is required
to ensure adequate hepatic regeneration and function following resection.
Patients with underlying hepatic dysfunction require a larger FLR: Patients
Hypertrophied
Left Liver
Embolized
Right Portal Vein
FIGURE 15.1 CT scan of the liver following PVE of the right portal vein; note the hypertro-
phy of the left liver.
0002086382.INDD 181 7/14/2014 4:25:54 PM

with steatosis or steatohepatitis require an FLR of 30% to 40%, and patients

with cirrhosis require a 40% to 50% FLR. FLR can accurately be measured
on CT or MRI using volumetric assessment. This is the most commonly
used method for measuring FLR in Western centers and is used to iden-
tify patients who may benefit from preoperative PVE. Indocyanine green
retention at 15 minutes has also been demonstrated to predict postopera-
tive liver failure and mortality; however, the availability of this technique in
Western centers is limited. Preoperative PVE is a technique that involves
embolization of the portal vein with cyanoacrylate, gelatin microspheres,
polyvinyl alcohol, or coils. Embolization of the tumor-bearing liver induces
hypertrophy of the contralateral liver and thus increases the FLR volume.
PVE is well tolerated with a low complication rate (< 5%) and normally
results in an absolute increase in FLR of approximately 8% to 16% depend-
ing on the degree of underlying hepatic dysfunction. The timing and degree
of hypertrophy in response to PVE may help guide patient selection for
operation.
A two-stage hepatectomy is a surgical approach that may be used in
patients with extensive disease involving both sides of the liver (Fig. 15.2).
In a subset of patients, clearance of all disease is often not possible in one
operation due to concerns of an inadequate FLR. When the CRLM are bilat-
eral, PVE is often not ideal as embolization of the entire tumor-bearing liver
is not feasible. In these circumstances, a two-stage approach may be pos-
sible. With the two-stage approach, the disease in one hemiliver is cleared
and the contralateral portal vein is occluded either by portal vein ligation
during this first operation or by PVE performed postoperatively. The rem-
nant portion of the liver that has been “cleared” is then allowed to hyper-
trophy, and then the patient is brought back to the operating room 4 to
6 weeks later to resect the remaining disease in the liver. When performing
a two-stage hepatectomy, most surgeons advocate initial resection of minor
2nd 1st
FIGURE 15.2 Depiction of a two-stage hepatectomy. The first stage was a segmentectomy
of the left lateral bisector; following PVE, the patient was brought back to the operating
room 6 weeks later for the second stage, which was a right hepatectomy.
002086382.INDD 182 7/14/2014 4:25:54 PM

disease (three or fewer segments) followed by resection of major disease

(greater than three segments) at the second operation. Doing the minor
liver resection first will spare a subset of patients the morbidity of a major
hepatic resection, as 20% to 30% of patients will never undergo the second
stage due to disease progression or worsening performance status. Of note,
patients undergoing a two-stage approach can safely be treated with che-
motherapy between stages without impairing hypertrophy of the FLR.
Improvements in chemotherapy for patients with CRC have substan-
tially improved tumor response rates in patients with CRLM (Fig. 15.3).
Indeed, 12% to 33% of patients with initially unresectable disease can be
downsized after receiving “conversion chemotherapy” to a point where
curative intent resection is feasible. Average 5-year survival for these
patients is 30% to 35%, which is significantly better than patients receiv-
ing chemotherapy alone (approximately 10%). In this subset of patients,
multidisciplinary involvement including close collaboration between the
medical oncologist and the surgeon is essential. Chemotherapy should be
continued until the patient is resectable, not until maximal response is
achieved. Prolonged chemotherapy beyond the point of resectability may
increase rates of chemotherapy-associated hepatotoxicity and contribute
to increased patient morbidity after hepatic resection. While the care of
each patient must be individualized, in general, preoperative cytotoxic che-
motherapy (e.g., FOLFOX or FOLFIRI) should be stopped about 4 to 6 weeks
prior to surgery, while bevacizumab (Avastin) should generally be stopped
6 to 8 weeks prior to surgery.
Ablation can involve either radiofrequency ablation (RFA) or micro-
wave ablation. Ablation either alone or in combination with resection
can increase the number of patients eligible for liver-directed therapy for
colorectal metastases. Ablation should not be viewed as a substitute for
hepatic resection but as a complementary therapy in patients in whom
complete resection is not possible. Often in patients with extensive disease,
FIGURE 15.3 CT scan of a patient pre- and postreceipt of preoperative chemotherapy.

Note the cytoreduction of the tumor and the decrease in the size of the masses following
chemotherapy.
0002086382.INDD 183 7/14/2014 4:25:56 PM

larger lesions can be resected, while smaller lesions are ablated. Ablation
can be performed percutaneously, laparoscopically, or during laparotomy
and can be used to successfully ablate tumors of up to 3 to 4 cm in diameter
with a single application. Generally, a 1-cm “margin” of thermal necrosis is
desired around the lesions. The efficacy of RFA in lesions adjacent to large
blood vessels decreases due to a heat-sink effect, which is associated with
higher local recurrence rates; microwave ablation may be less susceptible
to this “heat-sink” effect. Ablation should generally be avoided with lesions
near the hilum of the liver, as ablation may damage biliary structures and
lead to subsequent stricturing. The combination of ablation with hepatic
resection results in morbidity and mortality rates similar to those seen with
resection alone. It is difficult to compare survival of patients undergoing
ablation to those undergoing resection because patients who undergo abla-
tion often have more extensive disease and medical comorbidities than
those undergoing resection. Nonetheless, ablation can have an important
role in the management of patients with hepatic metastases as an adjunct
to the treatment plan in patients with unresectable disease.
Operative Strategy
At the time of hepatectomy, a thorough assessment for the extent of disease
including full liver mobilization, visual inspection, palpation, and intraopera-
tive ultrasound (IOUS) should be performed in all patients. IOUS is an impor-
tant tool to identify lesions intraoperatively and to define the relationship of
hepatic lesions to vascular structures. In several studies, IOUS has been dem-
onstrated to be superior to intraoperative inspection and palpation for iden-
tification of hepatic neoplasms. A systematic approach should be used to
perform IOUS in order to identify all hyper-, iso-, and hypoechoic lesions. All
suspicious lesions should be scanned in the transverse and sagittal planes to
define size and anatomic relationships. The IOUS characteristics of hepatic
metastases tend to be similar within patients (as compared to between
patients), and the echogenic appearance of the index lesion can be used to
predict the appearance of additional lesions in the same patient, facilitating
intraoperative identification of lesions. IOUS has been demonstrated in the
literature to change the operative plan in up to 67% of patients undergoing
hepatectomy. In addition, in 10% to 12% of patients, IOUS identifies at least
one lesion not seen on preoperative imaging when routinely employed. IOUS
should therefore be used in all patients undergoing surgery for CRLM to facil-
itate identification of all lesions and complete clearance of the liver.
Complete resection of CRLM with a microscopic negative margin
should be the goal when undertaking hepatectomy for CRLM. Numerous
studies have documented both a higher risk of recurrence and decreased
overall survival in patients with positive resection margins. Historically, a
1-cm resection margin was considered mandatory based on data from a
1999 study by Cady et al., which demonstrated improved overall survival in
patients with a 1-cm margin. A number of studies have subsequently dem-
onstrated no difference in local recurrence or overall survival in patients
with a subcentimeter R0 resection compared with those with margins
greater than 1 cm. In a 2008 study, de Hass et al. demonstrated that while
patients undergoing an R1 resection had increased risk of local recurrence,
there was no difference in overall survival when compared with patients
undergoing R0 resection. The authors attribute this change in outcome
after R1 resection for CRLM to increasingly effective chemotherapeutic
regimens. Thus, while the goal of surgery for CRLM remains an R0 resection
and the surgeon should strive for at least a 1-cm margin, resection should
be considered in patients where less than a 1-cm margin is obtainable as
long as all macroscopic disease can be resected.
002086382.INDD 184 7/14/2014 4:25:56 PM

TABLE
Outcomes Following Resection for Colorectal Liver
15.2 Metastases
5-Year
Number of Number of Patients Overall
Author Year Patients with R0 Resection Survival
Fong 1999 1,001 895 37%
Choti 2002 133 NS 58%
Abdalla 2004 190 NS 58%
Fernandez 2004 100 100 58%
Wei 2006 423 407 47%
Rees 2008 929 833 36%
de Jong 2009 1,669 1,391 47%
Morris 2010 3,116 NS 44.2%
NS, not specified.
OUTCOMES/FOLLOW-UP
The 5-year survival after curative intent hepatic resection for CRLM ranges
from 40% to 58% (Table 15.2). The NCCN currently recommends surveil-
lance with history, physical exam, CEA, and CT chest/abdomen/pelvis
every 3 to 6 months for 2 years and then every 6 months for years 3 to 5.
Over 60% of patients with resected hepatic disease will recur at some point
during their disease course. If patients do recur with liver-only disease,
repeat hepatectomy may be considered in a select group of patients. Several
studies have demonstrated perioperative morbidity and mortality of repeat
resection for CRLM similar to that for patients undergoing initial surgery.
Five-year overall survival for patients undergoing hepatectomy range from
29% to 42%; thus, an aggressive approach to repeat resection is warranted
in well-selected patients, as it may be the only chance for long-term cure.
CONCLUSION
Significant improvements in chemotherapeutic regimens, surgical tech-
nique, and patient selection have resulted in improved survival for patients
undergoing hepatic resection for CRLM. Many patients with disease that
was once considered unresectable are now able to enjoy the survival ben-
efits of resection for CRLM due to expanded indications for hepatectomy,
use of conversion chemotherapy, utilization of techniques to increase the
FLR, and liver-directed therapies. Careful preoperative planning and mul-
tidisciplinary collaboration are essential to ensure optimal outcomes for
these patients.
Suggested Readings
Adam R, Avisar E, Ariche A, et al. Five-year survival following hepatic resec-
tion after neoadjuvant therapy for nonresectable colorectal. Ann Surg Oncol
2001;8:347–353.
Asiyananbola B, Chang D, Gleisner AL, et al. Operative mortality after hepatic
resection: are literature-based rates broadly applicable? J Gastrointest Surg
2008;12:842–851.
Charnsangavej C, Clary B, Fong Y, et al. Selection of patients for resection of
hepatic colorectal metastases: expert consensus statement. Ann Surg Oncol
2006;13:1261–1268.
de Haas RJ, Wicherts DA, Flores E, et al. R1 resection by necessity for colorectal liver
metastases: is it still a contraindication to surgery? Ann Surg 2008;248:626–637.
0002086382.INDD 185 7/14/2014 4:25:56 PM

Fong Y, Fortner J, Sun RL, et al. Clinical score for predicting recurrence after hepatic
resection for metastatic colorectal cancer: analysis of 1001 consecutive cases. Ann
Surg 1999;230:309–318.
Mayo S, Pawlik T. Current management of colorectal hepatic metastases. Expert Rev
Gastroenterol Hepatol 2009;3:131–144.
Pawlik TM, Abdalla EK, Ellis LM, et al. Debunking dogma: surgery for four or more
colorectal liver metastases is justified. J Gastrointest Surg 2006;10:240–248.
002086382.INDD 186 7/14/2014 4:25:56 PM

16 Neuroendocrine and
Noncolorectal Liver
Metastases
Mark J. Truty and David M. Nagorney
NEUROENDOCRINE LIVER METASTASES

Introduction
Arguably, one of the greatest advances in solid tumor oncology in the
early twentieth century has been the evolving treatment of liver metasta-
ses. Improvements in modern imaging modalities, refinement in surgical
indications and techniques, more effective systemic therapies, and advanc-
ing technology in locoregional nonresectional strategies led to an increas-
ing number of treatment options for patients with hepatic metastases.
Although typically regarded as slow-growing neoplasms, neuroendocrine
tumors vary widely in their clinical behavior and biologic activity, and the
liver is the most common site of distant metastases. Various classification
systems and nomenclature exist for the clinical staging and management
of patients with primary neuroendocrine carcinoma (NEC). Herein, we will
overview the management of hepatic metastases from well-differentiated
NEC from various sites and summarize management options to provide
an algorithm and overall treatment strategy to approach these complex
patients. Importantly, although cure is rare, treatment typically leads to
prolonged survival, significant palliation of symptoms, and improved qual-
ity of life (QOL) for patients with metastatic NEC.
After colorectal cancer, NEC is the second most common indication
for liver-directed therapies in patients with metastatic disease. The relatively
indolent growth of these lesions compared to hepatic metastases from other
sites allows for a planned aggressive approach incorporating resection as a
major component of treatment. Surgical treatment must balance the signifi-
cant risk of treatment-related morbidity particularly if major hepatic or mul-
tivisceral resections are performed concomitantly. Although the operative
morbidity and mortality of such resections may be significant, specifically
with combined procedures, operation generally provides durable oncologic
or symptomatic benefits and rarely precludes medical therapy.
NECs of the gastrointestinal tract and pancreas are rare and comprise
less than 1% of all malignant disease. Recent prospective national tumor
database studies suggest a linearly increasing overall incidence of NEC, with
the second highest prevalence of all gastrointestinal cancers. Overall progno-
sis is dependent on the primary tumor type, site of origin, stage, and various
pathologic features such as size, histologic differentiation, and prolifera-
tive activity. No single system of nomenclature and grading for all NECs has
been universally accepted. Importantly, poorly differentiated or high-grade
NECs are more frequently associated with distant metastases (liver and bone
approximately 50% of cases) at initial diagnosis, and overall survival is poor
regardless of therapy. Thus, these patients are rarely candidates for hepatic
resection because of biologic aggressiveness. Overall, the therapeutic modali-
ties for management of patients with hepatic metastases from NEC include
187
0002086383.INDD 187 7/14/2014 7:09:41 PM

surgical, chemo-/biotherapy, ablative, and nuclear medicine strategies. Given

the heterogeneity of NEC, variable disease extent, and biologic behavior, the
recommendations for therapy are based on the results of retrospective out-
comes with sparse prospective high-level evidence.
Hepatic metastases from NEC are diagnosed either at the time of pri-
mary disease diagnosis (synchronous) or with disease progression (meta-
chronous). In general, the primary and regional extent of the NEC should
be resected before addressing hepatic metastases. Concomitant resection
of the primary NEC and the hepatic metastases should be undertaken only
after multidisciplinary assessment of the patient’s disease stage. Staged
resection of the primary NEC with subsequent hepatic resection of the
metastases is performed typically, although data from referral centers
suggest that concomitant resection can be performed safely in selected
patients. Most frequently, concomitant resections are undertaken in
patients requiring minor hepatic resections. In patients with clinical endo-
crinopathies, preoperative control is recommended. Preoperative octreo-
tide therapy is essential in patients with the carcinoid syndrome to prevent
intraoperative carcinoid crises (Fig. 16.1). Patients with carcinoid heart
disease occasionally require cardiac surgery to address valvular right heart
failure prior to hepatic resection (Fig. 16.2). Additionally, interval hepatic
arterial embolization in selected patients with bulky functioning NEC
metastases may be required in preparation for hepatic resection either to
address residual endocrinopathies after resection of the primary NEC or to
downsize intrahepatic metastases. Patients with pancreatic NEC generally
should undergo staged resections of the primary and hepatic metastases to
avoid excess morbidity from combined hepatic and pancreatic resection.
Diagnostic Evaluation
A comprehensive evaluation of patients with neuroendocrine liver metas-
tases includes identification of the anatomic site of origin (whether in
situ or previously resected), clinical and biochemical functional status,
FIGURE 16.1 Frequency of various symptoms in patients with the carcinoid syndrome.
002086383.INDD 188 7/14/2014 7:09:42 PM

Chapter 16 / Neuroendocrine and Noncolorectal Liver Metastases 189
tumor histopathology and grade, and evaluation of extent and distribu-

tion of metastatic disease for resectability. Tumor site of origin is usually
determined by endoscopy, contrast-enhanced enterography, radiolabeled
scintigraphy, or cross-sectional imaging computed tomography (CT) or
magnetic resonance imaging (MRI). NECs are termed functional if asso-
ciated with a discrete clinical syndrome. The syndrome is confirmed bio-
chemically through serum assays of respective neuroendocrine peptides.
The clinical neuroendocrine syndrome is specific to the dominant func-
tioning peptide. Not infrequently, multiple peptide levels are elevated
but rarely do patients present with a combined clinical endocrinopathy.
Nearly all NECs secrete chromogranin A; baseline concentration of this
marker should be obtained regardless of clinical syndrome. Other s ecretory
FIGURE 16.2 A. Contrast-enhanced CT with large right hepatic ileal carcinoid metas-
tasis. The inferior vena cava is dilated, and the liver has a heterogenous appearance
due to preexisting severe tricuspid valve regurgitation. B. Gross specimen of resected
tumor—the patient underwent tricuspid/pulmonary valve replacement prior to liver
resection.
(Continued )
0002086383.INDD 189 7/14/2014 7:09:43 PM

FIGURE 16.2 (Continued) C. Contrast-enhanced CT with low-density jejunal carcinoid

metastasis in left lateral sector. D. Gross specimen of resected liver tumor.
eptides should be e valuated as clinically indicated. Pathologic confirma-

p
tion of NEC by fine needle aspiration or core biopsy (either metastases
or primary tumor) is essential for clinical management. Grade of NEC is
defined by proliferative activity via the Ki-67 index and mitotic rate. These
key features of grading are critically important for clinical management.
Indeed, pathologic findings correlate, in part, with the malignant behavior
of NEC. In general, patients with well-differentiated grade 1 NEC benefit the
most from aggressive therapy because of typically indolent growth. Patients
002086383.INDD 190 7/14/2014 7:09:44 PM

with moderately differentiated grade 2 NEC have a less predictable and

more aggressive course though are often still candidates for operative treat-
ment. Operative intervention in patients with poorly differentiated grade
3 NEC is rarely indicated as these NECs are typically highly aggressive and
are characterized by rapid dissemination and resistance to most therapeu-
tic interventions. Currently, NEC is most frequently staged by the AJCC
TNM staging system. Staging is crucial for planning overall management.
As patients with hepatic metastases harbor by definition stage IV disease,
the site and extent of local–regional disease and the site, extent, and dis-
tribution of other distant metastases to the peritoneum, lung, and bone
significantly influence treatment planning.
Imaging
Radiolabeled somatostatin receptor scintigraphy (SRS) is the most sensitive
and accurate imaging modality for diagnosis and initial staging particularly
for identification of occult extrahepatic disease. Positive scintigraphy is
dependent upon the presence of active somatostatin receptors (SR) on the
NEC (Fig. 16.3). Current agents do not bind to all SRS receptors nor do all
NEC harbor all SRs. In addition, SRS also predicts response to somatostatin
analog therapy. More recently, fludeoxyglucose positron emission tomog-
raphy (18-FDG-PET) also has been used as an alternative staging adjunct.
However, contrast-enhanced CT remains the primary imaging study for
assessing resectability of hepatic metastases from NEC. Metastases are
hypervascular and are best visualized during the arterial phase of contrast
administration. In patients with iodinated contrast allergies or significant
steatosis, gadolinium-enhanced MRI is indicated; the contrast-enhanced
T2-weighed sequences are most sensitive (Fig. 16.4). In our experience, MRI
is complementary to CT with particular utility in detection of very small
hepatic metastases that may alter treatment strategies. Liver-specific con-
trast agents such as gadoxetate disodium (Eovist) increase the sensitivity
for detecting small NEC hepatic metastases. Such agents preferentially are
taken up by hepatocytes; thus, small contrast-negative metastases not visi-
ble on CT are more readily identified. Additionally, contrast-enhanced cross-
sectional imaging is essential to assess resectability of the primary NEC, if
present, and to assess residual or recurrent regional disease if previously
FIGURE 16.3 A. Computed tomography revealing numerous enhancing metastases in the

liver.
(Continued )
0002086383.INDD 191 7/14/2014 7:09:44 PM

FIGURE 16.3 (Continued) B. In-111 octreotide scan revealing multiple large foci of
intense radiotracer uptake in the liver as well as pancreatic tail, confirming metastatic
(arrow) pancreatic NEC. C. Gross specimen of resected liver tumors.
resected. Most patients with hepatic metastases from NEC harbor multifo-
cal bilobar disease, though 25% of patients have isolated lobar disease.
In potentially resectable patients, radiologic data should estimate the
future liver remnant volume in patients considered for extended or complex
hepatectomy or in patients with bulky parenchymal disease. Formal upper
and lower endoscopy and endoscopic ultrasound are also important tools
with which to identify primary NECs. Patients with carcinoid syndrome
002086383.INDD 192 7/14/2014 7:09:45 PM

FIGURE 16.4 Neuroendocrine liver metastases. A. Enhancing right lobe mass on the arte-
rial phase of computed tomography. B. Diffuse bilobar metastases on T2-weighted MRI
imaging.
should have a formal cardiac evaluation including e chocardiography to

evaluate the extent of right heart valvular disease. If present, valvular
disease should be corrected prior to any major liver resection.
Potentially Curative Resection

Although NEC is considered indolent, metastatic disease can be associ-
ated with severe, life-threatening endocrinopathies or symptoms from
locally invasive nonfunctional NEC. The leading cause of mortality aside
from hormonal complications is extensive hepatic progression and hepatic
failure. The timing and type of treatment should be considered in a multi-
disciplinary setting. Ideally, the first-line treatment for well-differentiated
hepatic metastases from NEC without extrahepatic spread and unilobar
alone or unilobar and limited contralobar metastases is resection with or
without concomitant ablation. Potentially curative resection (R0) is defined
as resection/ablation of all gross disease with an adequate functional liver
remnant. Fewer than 50% of patients undergoing hepatic resection of meta-
static NEC with hepatic metastases have R0 resection due to the extent of
0002086383.INDD 193 7/14/2014 7:09:45 PM

hepatic metastases. The historical 5-year survival for unresected patients

with NEC liver metastases is approximately 30%. In contrast, the expected
5-year survival for patients after R0 resection of hepatic metastases based
on retrospective series ranges from 60% to 80%. Factors correlating with
long-term survival include size and site of metastases, tumor grade, extent
of hepatic resection and baseline hepatic disease, presentation with stage
IV disease, and impact of endocrinopathy such as carcinoid heart disease.
Importantly, regardless of the apparent completeness of resection, disease
progresses in most patients, either extra- or intrahepatically. Most patients
will manifest disease progression within 5 years; the median identifica-
tion of disease progression is 18 months. Perioperative octreotide is rec-
ommended to prevent the carcinoid crisis in patients with the carcinoid
syndrome and other patients with demonstrable clinical control of endo-
crinopathies. Additionally, cholecystectomy is performed routinely during
hepatic resection to avert cholelithiasis associated with somatostatin ana-
log treatment and to potentially avoid ischemic cholecystitis or gallbladder
necrosis associated with future arterial embolization and ablation, respec-
tively. Adjuvant systemic chemotherapy is not established as the standard
of care, but participation in clinical trials where available is recommended.
Palliative Resection
QOL related to endocrinopathies can be improved significantly through
appropriate hepatic resection provided that residual disease is minimized.
In general, durable relief of endocrine symptoms is obtained when nearly all
radiographic metastases are resected even in the presence of gross residual
disease. Palliative cytoreductive resection should be considered primarily
in patients with endocrinopathies, particularly in patients who have failed
medical management. Whether survival in patients without endocrinopa-
thies is improved by palliative resection remains unproven. The complete-
ness and duration of response to palliative resection are primarily based
on residual metastatic volume and not resected metastatic volume. Given
the growth kinetics of NEC, at least 90% or more of the hepatic metastases
should be resected. In general, R2 resections are single-stage operations
and are parenchymal sparing because disease progression is inevitable.
Resection of the primary NEC may be undertaken concomitantly or at a
staged procedure (Fig. 16.5). Palliative resection has limited applicability
in patients with high-grade NEC or those with unresectable extrahepatic
FIGURE 16.5 Synchronous carcinoid metastasis. A. Computed tomography of large atypi-

cal carcinoid metastases replacing the left lateral sector.
(Continued )
002086383.INDD 194 7/14/2014 7:09:46 PM

FIGURE 16.5 (Continued) B. Gross intraoperative photo of this lesion. C. CT with infiltra-
tive primary carcinoid arising in the terminal ileum and involving the small bowel mesentery
(arrow). D. Gross intraoperative photo of this ileum tumor and the sclerotic adjacent mesentery.
metastases. Importantly, no high-level evidence has shown that debulking

improves survival or QOL compared to nonsurgical therapies. Palliative
resection can be enhanced by subsequent percutaneous ablative thera-
pies after hepatic regeneration. Notably, palliative resection does not
0002086383.INDD 195 7/14/2014 7:09:47 PM

reclude subsequent liver-directed therapies, antihormonal treatment, or

p
chemotherapy.
Liver Transplantation
The role of liver transplantation for metastatic NEC is limited. As may be
expected, complete resolution of endocrine symptoms is achieved after
liver transplant, but disease progression is typical, and survival has not
exceeded that of hepatic resection. Liver transplantation in patients har-
boring completely unresectable hepatic metastases who have controlled
primary and regional disease is clearly reasonable; however, selection cri-
teria are currently unclear. Patients who are young, have had resection of
the primary NEC without progression of extrahepatic disease for at least
6 months, and have low-grade NEC may be potential transplant candi-
dates (Fig. 16.6). Fewer than 1% of liver transplants in the United States are
performed for metastatic NEC.
FIGURE 16.6 A. Contrast-enhanced CT with numerous bilobar ileal carcinoid metastases

in a 26-year-old female. B. Intraoperative photos of liver lesions at the time of staged
primary tumor resection.
(Continued )
002086383.INDD 196 7/14/2014 7:09:48 PM

FIGURE 16.6 (Continued) C. Intraoperative photo of resected ileal carcinoid and mesen-
tery during staged primary tumor resection. D. Gross photo of liver explant following liver
transplantation 18 months after primary staged tumor resection and no evidence of interval
extrahepatic disease. E. Computed tomography 3 years following transplantation showing
no evidence of recurrent disease.
0002086383.INDD 197 7/14/2014 7:09:49 PM

Ablative Therapy
Ablation as a single therapy is applicable in patients with limited unresect-
able hepatic NEC metastases. Ablation is most frequently employed as an
adjunct during resection or for progressive disease following resection.
Ablative techniques are either thermal—cryotherapy, radiofrequency abla-
tion (RFA), microwave, or chemical—alcohol or acetic acid. Metastatic NEC
is quite sensitive to ablative therapy. Ablation has been shown to effectively
control disease progression and palliate symptoms whether performed
during laparotomy, laparoscopically, or percutaneously with image guid-
ance. In general, ablation should be considered as an adjunct to hepatic
resection for deep-seated metastases during hepatic resection and as an
adjunct to R2 hepatic resection postoperatively to further reduce residual
disease. Ablation should be considered as a primary liver-directed therapy
in patients with limited hepatic disease progression after prior hepatic
resection and as the initial liver-directed therapy for limited deep small
metastases to avoid major hepatic resection at the initial presentation
of the NEC. Ablative therapy should be avoided in patients with numer-
ous metastases, large metastases (> 5 cm), or juxtahilar metastases or in
patients with prior bilioenteric anastomoses because of the risk of chronic
intrahepatic infection.
Transarterial Therapies
Because NECs are hypervascular, hepatic metastases from NEC are well
suited for transarterial embolization therapies. This locoregional treatment
should be considered in patients with hepatic dominant disease where
resection and ablation are not tenable. Transhepatic arterial embolization
(TAE) may either be “bland” or use particles loaded with various additional
chemotherapeutic agents (transarterial chemoembolization—TACE). Both
TAE and TACE effectively control symptoms for up to 1 year and result in
significant decrease in biochemical markers and objective tumor responses
in about 50% of patients. Response to TACE is greater for high-grade NEC;
however, treatment of high-grade tumors is associated with greater toxicity.
Selective peripheral percutaneous embolization has replaced proxi-
mal hepatic arterial embolization to minimize hepatic ischemia– and
procedure –related liver failure. Moreover, embolization is frequently
staged; the lobe harboring the dominant metastatic burden is embolized
initially. Both TAE and TACE should be performed in experienced centers
that have developed clear-cut treatment algorithms. The most common
side effect of TAE/TACE is postembolization syndrome (PES), an inflam-
matory reaction triggered by tumor necrosis and/or hepatocellular injury.
This reaction occurs in up to 80% of patients either immediately following
the procedure or up to 2 weeks after the procedure. Typically, PES includes
fever, pain, extreme fatigue and malaise, and nausea/vomiting. A chemical
and/or ischemic cholecystitis may also occur if the cystic artery is inad-
vertently embolized. Contraindications to TAE/TACE include ipsilateral or
complete portal vein thrombosis, hepatic insufficiency, and previous bil-
ioenteric anastomoses such as after pancreaticoduodenectomy. Although
embolization can be repeated, eventually the arterial supply to residual
metastases becomes pruned with development of alternative arterial col-
laterals that become inaccessible to further embolization.
An alternative to TAE or TACE is selective internal radiation therapy
(SIRT). In SIRT, radioactive 90Yttrium is incorporated into microbeads that
are delivered transarterially to lodge within the neovasculature of the metas-
tases. This treatment delivers irradiation locally and concomitantly reduces
tumor blood supply. Although embolic in concept, the major arterial sup-
ply remains intact, which permits subsequent TAE/TACE. Candidates for
002086383.INDD 198 7/14/2014 7:09:49 PM

SIRT must harbor hepatic dominant disease and have angiographic proof
of limited extrahepatic arterial shunting. SIRTs should be considered in
patients with miliary hepatic metastases in particular. Repeat SIRT is pos-
sible. Current data suggest that SIRT is equally efficacious as TAE in achiev-
ing symptom control.
Radiation Therapy
External beam irradiation for the treatment for hepatic metastases has
been limited by the inability to deliver effective cytotoxic radiation doses
to tumors without associated significant parenchymal damage. Because
NECs have SRs that bind endogenous somatostatin, radiolabeled soma-
tostatin analogs potentially can deliver radioactive compounds directly to
the metastases. Peptide receptor radionuclide therapy (PRRT) combines
octreotide with various radionuclides. Radiopeptides are typically deliv-
ered through selective arterial perfusion as in SIRT. The radioparticles are
retained preferentially in the metastases due to selective receptor binding
delivering high-dose local irradiation while sparing normal liver tissue.
Radionuclides vary (90Yttrium, 111Indium, 177Lutetium); each emits radia-
tion of differing type and depth of tissue penetration. Although conceptu-
ally attractive, most agents are still investigational. Selective radiolabeled
therapies have shown significant benefits, specifically in patients with
recalcitrant tumor and symptoms after TAE/TACE. Objective response to
PRRT has been seen in up to 50% of patients. Furthermore, PRRT has been
shown to be effective treating the hypoglycemia of metastatic insulinoma
and is recommended after failure of standard treatments. The overall role
for radiolabeled receptor therapy in conjunction with other liver-directed
therapies including surgery is evolving.
Hormonal Therapy
Somatostatin analog treatment is the gold standard medical therapy for
patients with inoperable carcinoid cancer. In most patients with the car-
cinoid syndrome, control is often urgently required, and somatostatin
analogs provide prompt relief. Treatment is initiated with short-acting ana-
logs to confirm response because octreotide only binds to some SR. Once
response is confirmed, long-acting analogs are used to maintain response
with monthly injections. Randomized trial data support octreotide use for
control of symptoms and improved QOL. Additionally, octreotide has an
antiproliferative, apoptotic effect. Although used as a first-line treatment in
patients with advanced, unresectable midgut NEC, response in this setting
is frequently partial (< 10%) and typically stabilizing (50%). Other soma-
tostatin analogs that bind with higher affinity to various SRs are emerging
and will likely play a greater role in primary and adjuvant therapy in the
future. Interferon alpha has also been utilized however carries significant
and often poorly tolerated side effects. Somatostatin analogs are also typi-
cally used as adjuncts in combination therapy. Eventual treatment failures
occur after 1 to 2 years due to tachyphylaxis from the larger doses that are
required. Gastric antisecretory agents such as H2 blocker and PPI should
be used in all patients with gastrin-secreting tumors; similar cumulative
tachyphylaxis is also seen with these agents (Fig. 16.7).
Chemotherapy
The role of systemic chemotherapy for metastatic NEC is limited, though
chemotherapy efficacy has been seen in well-differentiated NEC. Regardless
of agents, response to chemotherapy is dependent on the site of NEC ori-
gin and tumor grade. Chemotherapy is considered second-line therapy
for midgut NEC because response rates are typically only 10% to 15%.
0002086383.INDD 199 7/14/2014 7:09:50 PM

In contrast, various combinations of streptozocin and doxorubicin/5-FU,

temozolomide and thalidomide, or single-agent dacarbazine may be con-
sidered in patients with advanced unresectable or progressive low-grade
metastatic pancreatic NEC. Objective response rates approach 30%, and
overall survival is improved compared to no therapy. More than 50% of
patients with poorly differentiated NEC (regardless of origin) respond to
combinations of etoposide and cisplatin. Systemic treatment failure in
FIGURE 16.7 A. Computed tomography of a hypodense duodenal gastrinoma metastasis.

B. In-111 octreotide scan revealing left medial sector gastrinoma metastasis with normal
physiologic uptake of indium-111 in the gallbladder, spleen, and kidneys.
(Continued )
002086383.INDD 200 7/14/2014 7:09:50 PM

FIGURE 16.7 (Continued) C. Gross specimen of resected liver tumor.
poorly differentiated NEC is typically associated with rapid progression and

death.
Numerous targeted agents have been used to treat patients with
metastatic NEC. Typically, NECs express vascular endothelial growth fac-
tor (VEGF) and its receptor (VEGFR) making them candidates for antian-
giogenic agents such as sunitinib (tyrosine kinase inhibitor with activity
against VEGFR) or bevacizumab (a VEGF antibody). Trials have demon-
strated either disease stability (in most patients) or objective partial regres-
sion (in a minority) with anti-VEGF therapy. The mTOR pathway is altered
in some NECs, particularly those pancreatic in origin. Everolimus, an orally
active mTOR inhibitor, has shown efficacy in low- to intermediate-grade
NEC as demonstrated in the multicenter RADIANT Trial. Everolimus has
been less effective in nonpancreatic NEC. Other agents (gefitinib, imatinib
temsirolimus, bortezomib, etc.) have been used to treat NEC patients, but
significant and durable tumor responses are rarely obtained (< 10%) with
targeted agents alone. Thus, targeted agents are not typically considered
first-line therapy and may be more effective as a chemotherapy adjunct.
Treatment Follow-up Evaluation

Current data show that despite surgical and medical therapy, disease pro-
gression occurs in more than 80% of patients with NEC hepatic metas-
tases. Given the high rate of recurrence following potentially curative
resection or progression following palliative procedures, continued sur-
veillance is warranted. Typically, follow-up is based on serial abdominal
imaging with CT or MRI and serum biomarkers (chromogranin A, neuron-
specific enolase, and other specific neuroendocrine peptides [if functional
NEC]) approximately every 3 to 6 months for low-grade NEC and every
2 to 3 months for high-grade NEC or in patients with documented symp-
tom recurrence or progression. Functional (SRS) whole body imaging is
performed only if the extent of the metastatic disease would affect liver-
directed therapy.
HEPATIC METASTASES FROM NONCOLORECTAL,

NONNEUROENDOCRINE MALIGNANCIES
As a common site for metastatic spread of most solid tumors, metastases
apparently confined to the liver will frequently arise from noncolorectal,
nonneuroendocrine malignancies. Given the benefits realized after resec-
tion of colorectal and neuroendocrine metastases, hepatic resection may
0002086383.INDD 201 7/14/2014 7:09:51 PM

be appropriate for some patients with noncolorectal, nonneuroendocrine

metastases. The low frequency of liver-limited disease and the wide
variation of underlying histopathology and tumor biology make outcome
prediction difficult in this patient group as a whole. In general, resection
of hepatic metastases from noncolorectal, nonneuroendocrine malignan-
cies may be considered if the primary and regional malignancy has been
resected, whole body imaging excludes extrahepatic metastases, and the
performance status of the patient permits resection. Resection of such
metastases should be undertaken only as one component of a multidisci-
plinary treatment approach with discussions among various cancer and
procedural specialists. Favorable selection factors for operation related
to the tumor include metachronous presentation of the hepatic metas-
tasis, long recurrence interval, number of metastases, extent of hepatic
resection, and documented responsiveness to chemotherapy. Clearly, the
potential survival benefit of resection must offset operative risk. The entire
supportive data for hepatic resection in such patients are based on limited
series of retrospective studies in highly selected patients. In general, resec-
tion of liver metastases from primary gastrointestinal tumors (esophageal,
pancreatic, gastric, and biliary) is unlikely to lead to meaningful survival
benefit. In contrast, resection of duodenal and small bowel liver metasta-
ses (more similar to colorectal cancer in biology) may have more favor-
able outcomes. Other highly selected patients with nongastrointestinal
liver metastases may potentially benefit from an aggressive approach. For
example, although less than 5% of metastatic breast cancer is limited to the
liver, the overall prevalence of the disease in the general population makes
this scenario relatively frequent. Patients with hormone receptor–positive
tumors have the best outcomes after attempts at curative liver resections;
however, most breast cancer liver metastases do not express receptors
(triple negative). Thus, the overall median survival after hepatectomy for
breast cancer metastases is typically 12 months, and development of extra-
hepatic disease is common after resection. Resection of hepatic renal cell
carcinoma metastases results in a good prognosis similar to outcomes for
colorectal metastases resection. The best outcomes in renal cell metastasis
patients occur in those with a prolonged disease-free interval. Resection of
sarcoma liver metastases remains the only potentially curative option given
the limited effective systemic therapies. Outcomes appear to be improving
for metastatic gastrointestinal stromal (GIST) tumors due to development
of effective targeted chemotherapy agent. Melanoma, specifically ocular in
origin, harbors one of the worst prognoses after liver resection due to either
occult or metachronous development of disseminated disease soon after
liver resection.
The standard approach to hepatic resection for malignancy should
be employed when considering atypical indications. Important planning
includes evaluating the margins of resection and adequate functional liver
remnant. The operative approach may be laparoscopic or open depending
on individual surgeons’ experience preference. Patients with noncolorectal,
nonneuroendocrine liver metastases may also be treated with percutane-
ous ablation. Some noncolorectal, nonneuroendocrine metastases such as
renal cell cancers or melanomas are quite vascular, and this feature may
predict responsive to transarterial therapy. Based on the experience of iso-
lated limb perfusion for melanoma and sarcoma, isolated hepatic perfusion
is a potential option at specialized centers for highly selected patients with
liver-only unresectable disease. This regional treatment allows higher doses
of chemotherapy to be delivered directly to the tumor, limiting systemic
toxicity.
002086383.INDD 202 7/14/2014 7:09:51 PM

Suggested Readings
Adam R, et al. Hepatic resection for noncolorectal nonendocrine liver metastases:
analysis of 1,452 patients and development of a prognostic model. Ann Surg
2006;244(4):524–535.
Ahmed A, et al. Midgut neuroendocrine tumours with liver metastases: results of the
UKINETS study. Endocr Relat Cancer 2009;16(3):885–894.
Ercolani G, et al. The role of liver resections for noncolorectal, nonneuroendocrine
metastases: experience with 142 observed cases. Ann Surg Oncol 2005;12(6):
459–466.
Glazer ES, et al. Long-term survival after surgical management of neuroendocrine
hepatic metastases. HPB (Oxford) 2010;12(6):427–433.
Gupta S, et al. Hepatic arterial embolization and chemoembolization for the treat-
ment of patients with metastatic neuroendocrine tumors: variables affecting
response rates and survival. Cancer 2005;104(8):1590–1602.
Gurusamy KS, et al. Palliative cytoreductive surgery versus other palliative treatments
in patients with unresectable liver metastases from gastro-entero-pancreatic neu-
roendocrine tumours. Cochrane Database Syst Rev 2009;(1):CD007118.
Gurusamy KS, et al. Liver resection versus other treatments for neuroendocrine
tumours in patients with resectable liver metastases. Cochrane Database Syst Rev
2009;(2):CD007060.
Knox CD, et al. Survival and functional quality of life after resection for hepatic carci-
noid metastasis. J Gastrointest Surg 2004;8(6):653–659.
Leblanc F, et al. Comparison of hepatic recurrences after resection or intraoperative
radiofrequency ablation indicated by size and topographical characteristics of the
metastases. Eur J Surg Oncol 2008;34(2):185–190.
Mayo SC, et al. Surgical management of hepatic neuroendocrine tumor metasta-
sis: results from an international multi-institutional analysis. Ann Surg Oncol
2010;17(12):3129–3136.
Pavel M, et al. ENETS Consensus Guidelines for the management of patients with liver
and other distant metastases from neuroendocrine neoplasms of foregut, midgut,
hindgut, and unknown primary. Neuroendocrinology 2012;95(2):157–176.
Que FG, et al. Hepatic resection for metastatic neuroendocrine carcinomas. Am J Surg
1995;169(1):36–42; discussion 42–43.
Reddy SK, et al. Resection of noncolorectal nonneuroendocrine liver metastases: a
comparative analysis. J Am Coll Surg 2007;204(3):372–382.
Sarmiento JM, et al. Surgical treatment of neuroendocrine metastases to the liver: a
plea for resection to increase survival. J Am Coll Surg 2003;197(1):29–37.
Strosberg J, Gardner N, Kvols L. Survival and prognostic factor analysis of 146 meta-
static neuroendocrine tumors of the mid-gut. Neuroendocrinology 2009;89(4):
471–476.
Touzios JG, et al. Neuroendocrine hepatic metastases: does aggressive management
improve survival? Ann Surg 2005;241(5): 776–783; discussion 783–785.
Vinik AL, Perry RR. Neoplasms of the gastroenteropancreatic endocrine system. In:
Holland JF, Bast RC, Morton DL, et al. (eds.), Cancer medicine. Baltimore, MD:
Williams & Wilkins, 1997:1605–1641.
0002086383.INDD 203 7/14/2014 7:09:51 PM

17 Benign Liver Tumors
Jason B. Liu and Marshall S. Baker
Liver lesions are being detected incidentally with increasing frequency on

imaging done as part of the evaluation of vague abdominal symptoms.
The primary question regarding these lesions is to distinguish benign
from malignant. Radiologic imaging broadly characterizes hepatic lesions
as solid or cystic. Patient history and radiographic patterns can further
aid in determining risk of malignancy and characterizing the lesion as
either benign or malignant. For instance, risk factors for significant
hepatic disease including a history of alcoholism, viral hepatitis, autoim-
mune hepatitis, primary biliary cirrhosis, hemochromatosis, hemosidero-
sis, or primary sclerosing cholangitis aid the clinician in assessing risk of
malignancy. Long-term oral contraceptive (OCP) or anabolic steroid use
confers increased risk of adenomatous change in the liver. This chapter
focuses on the presentation, diagnosis, and management of benign liver
lesions.
BENIGN CYSTIC LESIONS OF THE LIVER

Simple Cyst
Presentation
Simple hepatic cysts are found in 2.5% to 18% of the population. They may
be single or multiple. These cysts are serous fluid-filled sacs lined by cuboi-
dal biliary-type epithelium without communication to bile ducts. They are
typically less than 1 cm in diameter and are most commonly detected as
incidental findings on diagnostic imaging. Simple cysts tend to occur more
commonly in the right hepatic lobe and are more prevalent in women with
a peak in the fifth decade of life. Large cysts can be symptomatic due to
mass effect. Most commonly, large cysts will be associated with right upper
quadrant pain (Glisson capsule stretch), early satiety, or nausea. Rarely,
simple cysts will become large enough to cause compression of the vena
cava or biliary obstruction. Patients may also develop pain from hemor-
rhage into the cyst, bacterial superinfection of intracystic blood products,
or free rupture. Approximately 5% of patients will require intervention due
to symptoms.
Diagnosis
Ultrasonography is the most useful diagnostic test for simple cysts, which
appear as anechoic fluid-filled spaces without septations and with poste-
rior acoustic enhancement indicating a well-defined fluid–tissue interface.
Computed tomography (CT) imaging demonstrates a well-demarcated cystic
lesion with water attenuation and without contrast enhancement. Similar fea-
tures are seen on magnetic resonance imaging (MRI). Fine needle aspiration
is typically not required for diagnosis because the radiographic appearance is
definitive. Recently, microbubble contrast-enhanced ultrasound (CEUS) has
been used to confirm the absence of vascular flow within simple cysts.
204
002086384.INDD 204 7/15/2014 11:13:41 AM

Chapter 17 / Benign Liver Tumors 205
Treatment
Most simple cysts do not require treatment or follow-up. Giant simple cysts
(≥ 6 cm) that become symptomatic warrant operative management. Cysts
that are noted to be increasing in size raise concern for a cystadenoma or
cystadenocarcinoma and also warrant operative management. For patients
in whom imaging is not definitive or patient history raises suspicions for pos-
sible hydatid disease, Echinococcus should be ruled out with serology prior to
surgical intervention. Methods for treatment of simple cysts include percuta-
neous aspiration with or without a sclerosant (e.g., alcohol, minocycline, or
tetracycline), laparoscopic or open marsupialization (“unroofing”), or lapa-
roscopic or open resection. Simple aspiration typically leads to guaranteed
recurrence. The recurrence rate is moderately decreased with the addition of
a sclerosant following aspiration, but minor complications such as nausea,
vomiting, or increased pain can occur. Repeated treatments may also be nec-
essary for large cysts. Laparoscopic marsupialization with fulguration of the
cyst epithelium remains the treatment of choice for giant, symptomatic cysts
with a reported recurrence rate as low as 2% in published series. Recurrence is
associated with incomplete unroofing and location within the posterior liver
segments. An omental pedicle flap placed within the cyst bed may decrease
recurrence by adhesion formation. Formal cystectomy or segmentectomy has
also been performed but is associated with increased morbidity and mortality.
Polycystic Liver Disease

Presentation
Polycystic liver disease (PLD) occurs in association with autosomal dominant
polycystic kidney disease (ADPKD) or as an isolated form of autosomal domi-
nant polycystic liver disease (ADPLD). The development of polycystic liver
occurs over time in patients with ADPKD, while renal cysts are rarely asso-
ciated with isolated ADPLD. The increased prevalence of polycystic livers in
ADPKD is attributed to the improved life expectancy of these patients. Hepatic
involvement is also associated with female gender and severity of renal disease.
PLD is discovered commonly during the fourth or fifth decade of life.
Patients are typically asymptomatic, but symptoms associated with hepatic
enlargement can occur. Abdominal pain in the setting of PLD is usually asso-
ciated with cyst hemorrhage or infection. Infection of hepatic cysts carries a
2% mortality rate and should be treated aggressively with intravenous antibi-
otics and percutaneous drainage. Liver function is preserved, and laboratory
values typically only demonstrate a mild elevation in γ-glutamyl transpepti-
dase (GGT) and alkaline phosphatase when symptoms are present.
Diagnosis
The same modalities used to diagnose simple hepatic cysts are used to
diagnose PLD (Fig. 17.1). Ultrasound remains the preferred diagnostic tool
due to its accuracy and low cost. Three types of PLD have been suggested
based upon CT characterization of the number and size of cysts as well as
the amount of hepatic parenchyma between cysts. Type I patients have a
limited number of large cysts with large areas of normal hepatic paren-
chyma. Type II patients have multiple medium-sized cysts isolated to cer-
tain segments. Type III patients are characterized by near replacement of
the hepatic parenchyma with diffuse cysts. Surgical intervention for PLD is
grossly reflective of these three categories.
Treatment
Therapy for PLD is aimed at minimizing the proportion of hepatic cysts while
maximizing hepatic parenchyma, thereby relieving symptoms associated
0002086384.INDD 205 7/15/2014 11:13:41 AM

FIGURE 17.1 Axial CT imaging for a 54-year-old male presenting with epigastric pain. He
was found to have isolated PLD and underwent laparoscopic left hepatic lobectomy and
marsupialization of the dominant right hepatic cyst. He is currently asymptomatic 3 years
postresection.
with mass effect while avoiding the risk for liver failure. Strategies are iden-
tical to the management of simple hepatic cysts except for the addition of
orthotopic liver transplantation in diffuse type III disease. Patients with type
I disease are approached with marsupialization or percutaneous drainage
with sclerotherapy. No significant differences in morbidity, mortality, or
recurrence rates have been seen using laparoscopic versus open approaches.
Hepatic resection becomes applicable in patients with type II or type III dis-
ease whereby removal of multiple segments that are grossly affected can be
beneficial in volume reduction. Transplantation is reserved for patients who
have severely impaired quality of life, are in intractable pain and for whom
there are no other surgical options for management. Candidates usually have
cachexia, significant weight loss, malnutrition, recurrent cyst infections,
portal hypertension, and ascites. The decision to pursue liver transplanta-
tion requires weighing the high mortality rates (12% to 30%) associated with
transplant against the potential improvement in quality of life.
Cystadenoma
Presentation
Cystadenoma is the most common primary cystic neoplasm of the liver with
an associated malignant potential. The incidence is limited to case reports and
small case series. These cysts occur most commonly in women between 40 and
60 years of age, are discovered incidentally on diagnostic imaging, and are gen-
erally asymptomatic. Patients may develop symptoms of right upper quadrant
pain and anorexia due to mass effect from cyst growth.
Diagnosis
Diagnosis of cystadenoma is based primarily on ultrasonographic find-
ings. These include irregular borders, hypoechoic cyst with hyperechoic
002086384.INDD 206 7/15/2014 11:13:42 AM

septations, solid components, papillary projections, wall enhancement,

and dorsal shadows demonstrating calcified areas. Difficulty arises when
differentiating between cystadenoma, cystadenocarcinoma, complex cysts
(i.e., simple cysts with resolved internal hemorrhage), and hydatid cysts.
Multiphase CT or diffusion weighted MRI imaging may reveal vascularity
of the septa perhaps suggesting cystadenocarcinoma. Hydatid cysts dem-
onstrate similar patterns on imaging, and Echinococcus should be ruled out
with serology. CEUS can be helpful to rule in cystadenoma and cystadeno-
carcinoma as this modality may detect vascular flow, which is absent in
complex cysts. Analysis of cyst fluid nearly always shows elevated CA 19–9
concentration.
Treatment
Surgical resection with clear margins is the definitive treatment for cyst-
adenoma due to the risk of malignant transformation, which is generally
reported to be 10% to 15%. Some authorities consider enucleation effective
therapy for cystadenoma.
Hydatid (Echinococcus) Cyst: Cystic Echinococcosis and Alveolar

Echinococcosis
Hydatid cysts arise from infection with the tapeworm Echinoccocus gran-
ulosus that live in the small intestines of canines and sheep. Humans are
accidental intermediate hosts due to ingestion of eggs via the fecal–oral
route. Other virulent species include E. multilocularis, E. vogeli, and E. oli-
garthus. Echinococcal disease is endemic in many Mediterranean coun-
tries, Eastern Europe, the Middle East, East Africa, East Asia, Australia,
and South America. In the United States, most cases occur in immigrants
from endemic regions. Infection with E. granulosus causes cystic echi-
nococcosis (CE), while infection with E. multilocularis causes alveolar
echinococcosis (AE).
Cystic Echinococcosis (CE)

Presentation. E. granulosus has an annual incidence of 1 to 200 per 100,000
and is endemic to temperate climates such as the Mediterranean, Central
Asia, Australia, and South America where pastoral communities are promi-
nent. Cyst growth is slow and can remain asymptomatic for years. Cysts
can be found throughout the body with the liver (approximately 80%) most
commonly affected followed by the lung (approximately 20%). Symptoms
occur due to either mass effect (e.g., cholestasis, portal hypertension, Budd-
Chiari syndrome) or rupture. Severe presentations include bacterial super-
infection/hepatic abscess formation, secondary cholangitis due to rupture
into the biliary tree, or peritonitis and anaphylaxis due to intra-abdominal
rupture.
Diagnosis. CE is diagnosed based upon patient history, clinical findings

(e.g., abdominal pain, fever, chest pain, and dyspnea), ultrasonography,
and positive serology. Eosinophilia is rarely present unless there is already
leakage of antigen into the circulation. Ultrasonography has a sensitivity
of 90% to 95% and is the imaging modality of choice (Fig. 17.2). The World
Health Organization (WHO) classification bases treatment decisions on
ultrasound findings, which differentiate cysts into active, transitional, and
inactive states (Table 17.1). MRI or magnetic resonance cholangiopancrea-
tography (MRCP) can be used to evaluate any lesions that may involve the
biliary tree or for preoperative planning. The diagnosis is further corrob-
orated with detection of serum antibodies, which has a sensitivity range
between 85% and 98%.
0002086384.INDD 207 7/15/2014 11:13:42 AM

FIGURE 17.2 Computed tomography (A) and ultrasound (B) imaging of an echinococcal
cyst in a 74-year-old male of Greek origin. This cyst demonstrates ultrasound features typi-
cal of a CE4 type cyst. It is heterogeneous with degenerative contents and no daughter
cysts. On CT the wall is thickened and partially calcified. These findings are characteristic
of inactive cysts. The patient was managed with observation and has been asymptomatic
over 3 years of follow-up.
Treatment. There have been no randomized trials regarding the treatment

of CE. Treatment decisions are made based upon the standardized WHO
ultrasound-based classification as well as the available medical and surgical
expertise in the endemic region. Surgery, percutaneous treatments, and
antiparasitic drug therapy are the mainstays.
002086384.INDD 208 7/15/2014 11:13:42 AM

0002086384.INDD 209
TABLE
Classification of World Health Organization
17.1 Percutaneous Drug Suggested Initial
Phase Characteristics Surgery Therapy Therapy Therapy
CE1 Active Unilocular, simple cyst with visible cyst No Yes Yes PAIR + albendazole
wall and hydatid sand, “snowflake sign”
CE2 Active Multiseptated daughter cysts within Yes Yes Yes Other percutaneous therapy +
mother cyst; “wheel-like,” “rosette-like,” albendazole
“honeycomb-like” structures
CE3a Transitional Detached laminated membrane from cyst No Yes Yes PAIR + albendazole
wall, “water-lily” sign
CE3b Transitional Solid with daughter vesicles, “complex Yes Yes Yes Other percutaneous therapy +
mass” albendazole
CE4 Inactive Heterogeneous, degenerative contents, No No No None
“ball of wool” sign
CE5 Inactive Calcified cyst wall, cone-shaped shadows No No No None
7/15/2014 11:13:42 AM
Surgery is indicated for the treatment of WHO class CE2-CE3b (active

cysts with multiple daughter vesicles), single liver cysts situated superficially
in the liver, infected cysts when percutaneous therapy is unavailable, cysts
communicating with the biliary tree, and cysts compressing adjacent vital
organs. The primary goal is to remove the entirety of parasitic material while
avoiding spillage and secondary echinococcosis. Total cystectomy can be
performed using open or laparoscopic techniques although the frequency
of spillage has not been compared. Total cystectomy is ideally performed
in a “closed” manner whereby the cyst is removed by means of a partial
hepatectomy without opening the cyst. Cystectomy done “open” whereby
the cyst is injected with protoscolicidal agents (i.e., 20% hypertonic saline),
unroofed, and the pericystic tissue is excised has also been described but is
believed to carry increased risk of dissemination. Involvement of the bili-
ary tree is more commonly found in large cysts greater than 7.5 cm. Biliary
communication can be detected intraoperatively by using dye or fluoros-
copy, and the communication can be simply suture ligated. Perioperative
benzimidazoles (e.g., albendazole, mebendazole) are recommended to
minimize the risk of perioperative dissemination.
Percutaneous treatment including PAIR (puncture, aspiration, injec-
tion, reaspiration) and modified catheterization techniques is indicated for
inoperable patients, recurrent disease, or cysts that fail to respond to benz-
imidazoles only. Communication with the biliary tree must be ruled out
prior to performing PAIR to avoid the risk of chemically induced sclerosing
cholangitis. Success is associated with many of the types of CE except for
CE2 and CE3b, which tend to relapse after PAIR. These types are treated
with surgical resection, as discussed above, or with modified catheteriza-
tion techniques that involve large-bore catheters, sclerosing agents, and
curettage.
Cyst types CE4 and CE5 (inactive or heavily calcified cysts with no
daughter cysts) can be simply surveyed without treatment. These are gener-
ally of limited clinical significance and unlikely to cause biliary complica-
tions or symptoms.
Alveolar Echinococcosis (AE)

Presentation. Infection with E. multilocularis has an annual incidence of
0.03–1.2 per 100,000 and is isolated to regions of central and eastern Europe,
Russia, China, and northern Japan. It has been noted that the frequency of
cases in Western Europe has increased with the popularity of fox hunting.
Interestingly, AE affects the liver primarily and does not form cysts like CE.
The larvae invade locally or spread hematogenously akin to malignancy.
The disease is chronic with an average incubation period of 5 to 15 years.
Symptoms include cholestatic jaundice, abdominal pain, fatigue, weight
loss, and hepatomegaly.
Diagnosis. Diagnosis of AE is based on patient history, clinical presen-

tation, ultrasonography, and positive serology. Cross-sectional imaging
including CT and MRI/MRCP are used to further evaluate the local extent
of the disease particularly during preoperative planning. Classification of
AE is based upon the WHO PNM staging system analogous to the TNM
staging of malignancy.
Treatment. After diagnosis is confirmed and the extent of disease delin-

eated, the PNM staging system is used to direct therapy. Treatment involves
surgical resection, orthotopic liver transplantation, antiparasitic drug
therapy, or endoscopic or percutaneous interventions. The primary goal of
treatment is complete surgical excision whenever possible. Endoscopic or
002086384.INDD 210 7/15/2014 11:13:42 AM

percutaneous interventions should be employed as palliative procedures

when surgery is not feasible. Benzimidazoles are always indicated and, in
the setting of surgical resection, given as lifelong adjuvant therapy.
Liver transplantation can be considered in the setting of patients who
have severe liver insufficiency secondary to biliary cirrhosis or Budd-Chiari
syndrome, recurrent cholangitis, inability to perform radical liver resection
(i.e., lack of hepatic reserve), and the absence of extrahepatic disease. As of
2003, 45 liver transplantations have been performed for AE with variable
success.
Pyogenic Liver Abscess

Presentation
Pyogenic (bacterial and fungal) liver abscesses (PLA) were once complica-
tions from inadequately treated appendicitis, diverticulitis, or other intra-
abdominal infections. Biliary sources are now the primary causes of PLA.
The incidence is estimated at 20 cases per 100,000 hospital admissions with
the average patient between 50 and 60 years of age and of male predomi-
nance. The incidence of PLA is not associated with ethnicity or geographic
location.
Abscesses can be divided into six categories based upon the route of
infection listed in order of descending frequency: (1) biliary (60%), (2) cryp-
togenic (17%), (3) hepatic artery (10%), (4) portal vein (7%), (5) penetrat-
ing trauma (5%), and (6) direct extension (3%). Iatrogenic causes include
biliary–enteric anastomoses, liver-directed therapies for malignancy, and
transplantation. Escherichia coli and Klebsiella pneumoniae are the most
common pathogens followed by Staphylococcus aureus, Enterococcus spp.,
Streptococcus viridans, and Bacteroides. The incidence of Klebsiella as the
primary pathogen in liver abscesses has increased recently, particularly
in those with diabetes mellitus. PLA have become a unique complica-
tion associated with patients undergoing chemotherapy with hepatotoxic
drugs, such as oxaliplatin (sinusoidal obstruction syndrome) and irinote-
can (steatohepatitis), for treatment of hepatic metastases. Drug-induced
liver injury is postulated to cause increased susceptibility to liver abscess
formation.
Fever, right upper quadrant abdominal pain, nausea, vomiting, mal-
aise, chills, and weight loss are common presenting symptoms. Elevated
transaminases and/or obstructive jaundice can be present on laboratory
evaluation. Leukocytosis is frequently but not always present. Bacteremia
is present 50% to 95% of the time. Severe complications such as intraperi-
toneal or pericardial rupture, empyema, and broncho–pleural–hepatic
fistulae are rare.

Diagnosis
The diagnosis of PLA is dependent on imaging studies, where they can be
solitary or multiple. The right hepatic lobe is most often involved, although
a miliary presentation can occur. CT imaging with intravenous contrast is
the preferred modality with a diagnostic accuracy of 93% to 96%. Imaging
reveals a hypodense lesion, uni- or multiloculated, without contrast
enhancement surrounded by rim enhancement with or without the pres-
ence of gas (Fig. 17.3). Cross-sectional imaging can also demonstrate a likely
cause in approximately 70% of cases.
Treatment
Treatment primarily involves parenteral antibiotics and percutaneous
drainage, which has a success rate of 60% to 90%. Antibiotics are usually
0002086384.INDD 211 7/15/2014 11:13:42 AM

FIGURE 17.3 Axial CT imaging of a pyogenic liver abscess presenting in a 49-year-old male
alcoholic. The abscess presented in the context of an acute bout of alcohol-inducted hepatitis
associated with portal vein thrombosis and was managed by percutaneous d rainage and
antibiotic administration.
continued for 4 to 6 weeks and tailored to culture results. Repeated per-

cutaneous aspiration without drain placement can be considered in those
with solitary, unilocular abscesses 1 to 3 cm in diameter. No randomized
controlled trial has compared percutaneous aspiration versus drain place-
ment. Percutaneous drainage carries a 5% to 6% mortality rate.
Surgical drainage is reserved for patients who fail percutaneous drain-
age. In patients with a biliary source of pyogenic liver abscess, failure of
percutaneous drainage is associated with biliary fistula. In these patients,
endoscopic biliary decompression should be combined with percutane-
ous drainage to resolve the abscess without requiring surgical interven-
tion. Surgical drainage can be approached laparoscopically or with formal
hepatic resection. No studies have compared the effectiveness of these
approaches.
Amebic Liver Abscess

Presentation
Liver abscess is the most common extraintestinal site of amebic infection
but occurs in less than 1% of Entamoeba histolytica infections. E. histolyt-
ica is endemic to tropical and temperate climates, such as Mexico, India,
Indonesia, sub-Saharan Africa, and Central and South America. It most
commonly affects men between 20 and 40 years of age.
The disease is contracted via ingestion of contaminated water or
food, which typically results in dysentery. However, the trophozoite can
burrow through the colonic mucosa in certain hosts to enter the portal
circulation and form an abscess within the liver. Clinical presentation
usually occurs within 8 to 20 weeks. Signs and symptoms most commonly
include right upper quadrant abdominal pain, high fever (>38.5°C), and
tenderness to liver palpation. Amebic dysentery and liver abscess rarely
occur together.
Laboratory evaluation demonstrates leukocytosis without eosino-
philia, elevated alkaline phosphatase, and occasional transaminitis.
002086384.INDD 212 7/15/2014 11:13:42 AM

Diagnosis
Imaging and serology remain the cornerstones of accurate diagnosis.
Enzyme-linked immunoabsorbent assay (ELISA) has a sensitivity of
99% and a specificity of 90% for amebic liver abscess. This test should be
repeated in 7 days if the initial result is negative due to the delayed forma-
tion of antibodies. Furthermore, patients from endemic regions may have
positive serology due to previous infection. Serum antigen detection assays
can also be used to aid in diagnosis.
It is difficult to distinguish between pyogenic and amebic liver abscess
on imaging as they demonstrate similar features (e.g., rim enhancement).
Percutaneous aspiration of an amebic abscess typically yields an odorless,
reddish-brown “anchovy paste” fluid composed of necrotic hepatocytes.
Trophozoites are seen in less than 20% of aspirates. Superinfection of ame-
bic abscesses with enteric bacteria may occur.
Treatment
Amebic abscesses are treated with an amebicidal agent and a luminal agent.
Metronidazole for 7 to 10 days and either paromomycin or iodoquinol for
7 days provide adequate treatment. Percutaneous aspiration or surgical
drainage is rarely necessary in amebic liver abscesses except when abscesses
are greater than 6 cm. Intracavitary injection of metronidazole during per-
cutaneous aspiration is an effective strategy for larger abscesses (Table 17.2).
BENIGN SOLID LESIONS OF THE LIVER

Hemangioma
Presentation
Cavernous hepatic hemangiomas are the most common solid benign liver
tumor with an overall prevalence of 5% to 20%. Hemangiomas occur pre-
dominantly in females (2–6:1) between 30 and 50 years of age. They are
thought to arise from congenital hamartomas due to progressive ectasia. No
causal link to estrogen therapy has thus far been identified. Hemangiomas
are most often asymptomatic and are discovered incidentally on abdominal
imaging. Symptoms occur in approximately 12% of patients, usually due to
capsular stretch or compression of adjacent structures. Symptoms include
right upper quadrant pain, palpable midepigastric mass, nausea, early sati-
ety, or dyspepsia. Rarely, patients may present with Kasabach-Merritt syn-
drome, a consumptive coagulopathy manifesting with thombocytopenia,
hypofibrinogenemia, and systemic bleeding. Hemangiomas have no malig-
nant potential and rarely if ever rupture.
Diagnosis
Ultrasonography yields a sensitivity of 60% to 70% and specificity of 60%
to 90% for identification. Multiphase CT and MRI are more accurate than
is ultrasound. Peripheral nodular contrast enhancement with centripetal
filling is the classic presentation of hemangioma on multiphase imaging.
Magnetic resonance imaging has reported sensitivity and specificity of 95%
in identifying hemangiomas. Because of this, excellent accuracy angiog-
raphy or biopsy is no longer used to diagnose hemangiomas. Surveillance
imaging is generally not recommended unless there is uncertainty regard-
ing the diagnosis or symptoms develop.
Treatment
Most hemangiomas remain asymptomatic and stable over time. Intervention
to prevent future complications is not justified. Rare case reports of hem-
angioma rupture associated with large size or trauma are not enough
0002086384.INDD 213 7/15/2014 11:13:42 AM

TABLE
002086384.INDD 214
Management of Benign Cystic Liver Lesions
17.2 Risk Factors or Potential Malignant Initial Diagnostic
Associations Complications Potential Modality Initial Therapy Follow-up
Simple Cyst Age > 60 y Portal hypertension, hemorrhage, None Ultrasound None unless None required
Female gender biliary obstruction, bacterial symptomatic regardless of

superinfection, rupture intervention
Polycystic Liver Intracranial aneurysms, mitral Portal hypertension, hemorrhage, Rare Ultrasound None unless None unless
Disease valve prolapse, mitral valve biliary obstruction, bacterial symptomatic symptoms recur
regurgitation, pancreatic cysts, superinfection, rupture
renal cysts
Cystadenoma Female gender Obstructive jaundice, chronic Yes Ultrasound/MRI Surgical resection None
cholecystitis, cholelithiasis
Cystic Endemic regions Portal hypertension, cholestasis, None Ultrasound and Depends on WHO Depends on WHO
Echinococcus Budd-Chiari syndrome, serology classification classification
bacterial superinfection,
rupture, anaphylaxis
Alveolar Endemic regions Cholestasis None Ultrasound and Depends on WHO Depends on WHO
Echinococcus serology classification classification
Pyogenic Liver Iatrogenic (biliary-enteric Rupture, empyema, broncho– None Multiphase CT Antibiotics ± None unless
Abscess anastomosis, liver direct pleural–hepatic fistula, septic percutaneous symptoms recur
therapies, transplantation) shock drainage
Amebic Liver Endemic regions, Rupture, empyema, broncho– None Ultrasound and Pharmacologic None unless
Abscess immunocompromise, history of pleural–hepatic fistula, septic ELISA symptoms recur
splenectomy shock
7/15/2014 11:13:43 AM
to w arrant prophylactic surgery. Surgical resection can be considered in

patients with symptoms, in lesions in which malignancy cannot be excluded,
or to treat complications. Enucleation is the operation of choice due to a sig-
nificantly lower rate of perioperative morbidity compared to hepatic lobec-
tomy. Radiofrequency ablation (RFA) and cryotherapy have also been used
in select situations.
Focal Nodular Hyperplasia

Presentation
Focal nodular hyperplasia (FNH) is the second most common benign solid
hepatic lesion after hemangiomas. The reported incidence of FNH is 0.3% to
3%, though only 0.03% have clinical significance. They are associated with
female gender (8:1) and occur between 30 and 50 years of age. Concurrent
adenomas are diagnosed in up to 3.6% of patients and concurrent heman-
giomas in up to 23% of patients with FNH. FNH is not associated with OCPs
or pregnancy. Lesions are typically found incidentally, and symptoms are
associated with capsular stretch or compression of adjacent structures. No
malignant potential and no potential to rupture spontaneously have been
associated with FNH.
Diagnosis
The classic description of FNH is a lobular mass with homogeneous arte-
rial enhancement, radiating fibrous septa, and a central, nonenhancing scar.
This pattern is detected, however, in only 40% to 50% of cases. The diagnosis
of FNH must be differentiated from that of hepatocellular adenoma as these
two entities are managed differently. Multiphase MRI with liver-specific con-
trast agents has been shown to have a sensitivity and specificity of 97% and
100%, respectively, in differentiating FNH from adenoma. In addition, FNH
should also be distinguished from fibrolamellar hepatocellular carcinoma
(HCC), which also demonstrates a central scar. Characteristics of fibrola-
mellar HCC include large size greater than 10 cm, heterogeneous enhance-
ment, presence of calcifications, and hypointensity of the central scar on
T2-weighted MRI imaging compared to the hyperintensity seen with FNH.
Treatment
Symptoms or the inability to exclude malignancy are the indications for
intervention. Hepatic enucleation or lobectomy remains the procedure of
choice, although interventional procedures involving embolization and
RFA have also been used.
Hepatocellular Adenoma
Presentation
HCA is a benign neoplasm composed of hepatocyte proliferation. These
lesions occur in the population with a female:male ratio of 11:1, and usu-
ally present between 30 and 40 years of life. Risk factors for HCA develop-
ment include OCP use and pregnancy in females and anabolic steroid use in
males. A subset of patients with glycogen storage diseases type Ia and type III
can develop HCA, typically with a male:female 2:1 predominance. Patients
with HCA present with pain 25% to 50% of the time or are otherwise asymp-
tomatic with the lesion incidentally discovered on axial imaging. They are
isolated lesions located commonly within the right hepatic lobe. Liver
adenomatosis occurs when more than 10 lesions are present (see below).
Lesions greater than 5 cm, OCP use, and pregnancy are risk factors for
rupture or hemorrhage with the classic presentation of sudden-onset right
upper quadrant pain followed by hypotension. Rupture and hemorrhage are
0002086384.INDD 215 7/15/2014 11:13:43 AM

relatively rare events. HCAs carry a potential for malignant transformation

in up to 11% of lesions. Adenomas occurring in men, lesions greater than
5 cm, those associated with androgen steroid use, and those associated with
beta-catenin gene mutations are more likely to be malignant.
HCAs are classified into three histopathologic varieties: (1) Steatotic
HCA caused by hepatocyte nuclear factor 1a (HNF1α) mutation, (2) inflam-
matory HCA caused by IL-6 signal transducer (IL6ST) mutation, or (3)
HCAs associated with β-catenin mutation. Inflammatory HCA is the most
common variant representing 35% to 50% of HCAs. The specific subgroup
of HCA involving β-catenin mutations may have a greater risk for malig-
nant transformation compared to those with HNF1α or IL6ST mutations.
Further investigation is needed to determine the benefit of histopathologic
diagnosis. Core needle biopsy itself carries a risk of sampling error, bleed-
ing, and the theoretical risk of biopsy tract seeding.
Diagnosis
The confirmatory diagnosis of HCA is challenging because of its heteroge-
neous imaging features, particularly in the setting of necrosis, old or recent
hemorrhage, and calcific foci. Ultrasound has no utility in diagnosis of
HCAs unless contrast enhancement is used. Contrast enhanced ultrasound
reveals a centripetal enhancement during arterial phase compared to a cen-
trifugal filling pattern in FNH. Multiphase CT scan is more useful than ultra-
sound to diagnose HCA, but its utility lies in its ability to identify FNH and
hemangiomas to rule out HCA. Multiphase MRI with liver-specific contrast
agents provides the best definitive diagnosis and can be used to suggest the
subtype. Steatotic adenomas display diffuse signal dropout on T1-weighted
chemical shift sequence associated with their high fat density, and are mod-
erately enhancing in the arterial phase without persistent enhancement in
the portal venous and delayed phases. This pattern yields a sensitivity and
specificity of 86% and 100%, respectively. Inflammatory adenomas display
high-intensity signal on T2-weighted MRI and lack of signal dropout of fat
suppression images and have strong arterial enhancement with persistent
enhancement of portal venous and delayed phases. This pattern yields a
sensitivity and specificity of 85% and 87%, respectively. Unfortunately, no
particular MRI pattern has been associated with β-catenin mutated lesions,
which carry the greatest potential for malignant transformation.
Treatment
Patients who take OCPs or anabolic steroids should be advised to discontinue
these medications as this may lead to regression of the lesion. Treatment
modalities include RFA, transarterial embolization (TAE), and hepatic
resection. Hepatic resection, either anatomic segmentectomy/lobectomy
or enucleation, is preferred. To date, no clinical trial has compared RFA or
TAE against surgical resection. Surgical resection is indicated in all adeno-
mas occurring in men, those that cause symptoms or hemorrhage, and those
greater than 5 cm in women. Adenomas that have ruptured should be man-
aged initially with resuscitation followed immediately by TAE. Resection
during acute rupture has a reported mortality of 8% to 10%. In addition to
hemorrhage control, TAE has been shown to cause tumor shrinkage. Lesions
that are smaller than 5 cm in size should be surveyed for growth. A typi-
cal regimen includes ultrasound examination every 3 months for 6 months
and extending the duration thereafter if no growth or suspicious features
are detected. Patients who develop symptoms should be offered resection.
Women with adenomas greater than 5 cm in size are advised to
undergo resection prior to conception. Women who are diagnosed with
adenomas during pregnancy are advised to undergo elective resection
002086384.INDD 216 7/15/2014 11:13:43 AM

FIGURE 17.4 Axial CT imaging demonstrating a large hepatic adenoma in a 55-year-old

woman presenting with hypotension and anemia. The scan demonstrates evidence of hem-
orrhage into the tumor. This patient was managed with selective embolization to the right
hepatic lobe mass and interval right hepatic lobectomy.
uring the second trimester for lesions greater than 5 cm in size. Lesions
d
that are discovered in the third trimester should be monitored with serial
ultrasound exams as the rare risk of spontaneous rupture exists (Fig. 17.4).
Hepatic Adenomatosis
Presentation
Liver adenomatosis refers to the patient with arbitrarily defined greater
than 10 HCAs. The presentation is identical to that of isolated HCAs. It has
been argued that adenomatosis is not a separate entity from solitary HCAs.
Diagnosis
Histologic and radiographic features of adenomatosis are identical to that
of isolated HCA.
Treatment
The optimal mode of treatment has not been defined due to the rarity of
presentation and the difficulty of isolated lobectomy. Treatment options
include a combination of surgical resection, RFA, and TAE. Orthotopic liver
transplantation has been performed in rare cases. The number of adeno-
mas is not associated with increased risk of malignant transformation or
hemorrhage. Treatment should be aimed at the lesion greater than 5 cm in
size with close surveillance of the remaining lesions.
Nodular Regenerative Hyperplasia

Presentation
Nodular regenerative hyperplasia (NRH) occurs when normal hepatic
parenchyma is transformed into 1- to 3-mm regenerative nodules sepa-
rated by atrophic areas without the presence of perisinusoidal fibrosis as
seen in cirrhosis. This condition is thought to occur due to altered blood
0002086384.INDD 217 7/15/2014 11:13:43 AM

flow causing hepatocyte hyperplasia in regions with adequate blood flow

and atrophy in regions with ischemia. NRH is a rare condition with autopsy
studies demonstrating an overall incidence between 0.7% and 2.6%. Our
understanding of the condition is limited to case reports with only approxi-
mately 350 cases reported from 1975 to 2011. It most commonly presents
in middle age and increases in frequency with age. There is no association
with gender or ethnicity. It is felt that NRH occurs in response to other sys-
temic perturbation including immunologic, hematologic, medication-asso-
ciated, cardiac, or pulmonary conditions.
Patients with NRH are usually asymptomatic and detected on workup
of the underlying systemic condition. When symptoms do occur, they are
the result of portal hypertension. Differentiation should be made between
NRH and large regenerative nodules (LRN), which occur in the presence of
hepatic outflow obstruction, such as in Budd-Chiari syndrome.
Diagnosis
Suspicion for NRH usually occurs during the evaluation of portal hyperten-
sion especially in the patient without evidence of cirrhosis. Imaging studies
in the diagnosis of NRH cannot readily distinguish LRN from the regenera-
tive nodules of cirrhosis. Imaging subsequently is most useful for the exclu-
sion of other hepatic lesions. Definitive diagnosis depends on liver biopsy
with histopathologic analysis.
Treatment
Treatment of NRH is geared towards management of the underlying condi-
tion and the management of portal hypertension when it occurs. Standard
therapies for portal hypertension should be employed, including dietary
changes, medical management of ascites, endoscopic intervention for vari-
ces, and surgical shunting procedures or transjugular intrahepatic porto-
systemic shunts (TIPS). Orthotopic liver transplantation can be considered
in the patient with hepatic failure. Mortality from NRH is most commonly
due to variceal bleeding.
Angiomyolipoma
Presentation
Angiomyolipomas (AML) are rare, benign mesenchymal tumors most often
found in the kidney. Only approximately 300 cases of hepatic AML have been
reported between 1976 and 2012. These lesions are well-circumscribed, soft
masses composed of mature fat cells, blood vessels and smooth muscle.
Their pathogenesis is unknown, and lesions are usually incidentally iden-
tified on diagnostic imaging. Both renal and hepatic AML are associated
with tuberous sclerosis complex. Patients reporting symptoms usually have
large AML causing capsular stretch or compression of adjacent structures.
Spontaneous rupture has been reported rarely.
Diagnosis
The accurate diagnosis of AML depends upon the proportion of fat, smooth
muscle, and blood vessels within the lesion. Generally, lesions that contain
predominantly fat are more accurately diagnosed with imaging and those
lesions containing less fat are commonly mistaken for malignancy (e.g.,
HCC, metastasis, liposarcoma). In general MRI with hepatic-specific con-
trast agents remain the best imaging modality for diagnosis, though the
overall accuracy of MRI is still poor. In a recent series of 79 patients under-
going surgical resection for presumed AML, 48% were diagnosed incor-
rectly preoperatively with imaging or angiography.
002086384.INDD 218 7/15/2014 11:13:43 AM

0002086384.INDD 219
TABLE
Management of Benign Solid Liver Lesions
17.3 Risk Factors or Potential Malignant Initial Diagnostic
Associations Complications Potential Modality Initial Therapy Follow-up
Hemangioma Female gender Kasabach-Merritt None Multiphase CT None unless None unless
syndrome, symptomatic symptomatic
rupture
Focal Nodular Female gender None None Multiphase MRI None unless symptomatic None unless
Hyperplasia or cannot exclude symptoms recur
malignancy
Hepatocellular Female gender, OCPs, Rupture, Yes Multiphase MRI Surgical resection if Surveillance for
Adenoma pregnancy, glycogen hemorrhage > 5 cm growth
storage disorders,
anabolic steroids
Hepatic Same as hepatocellular adenoma
Adenomatosis
Nodular Regenerative Secondary to other Portal hypertension None Tissue biopsy Management of None unless
Hyperplasia systemic conditions underlying condition symptomatic
Angiomyolipoma Tuberous sclerosis Rupture rarely Rarely Multiphase MRI Surgical resection if Surveillance for
> 6 cm or cannot growth
exclude malignancy
7/15/2014 11:13:43 AM
Treatment
The preferred treatment of AML is controversial due to the rarity of
the c ondition. In general, AML has been treated as a benign liver lesion;
however, case reports of malignant transformation presenting as vascu-
lar invasion or metastatic disease have been published. Because of these
observations, surgical resection is recommended in lesions greater than
6 cm, lesions that demonstrate rapid growth during surveillance, or lesions
in which malignancy cannot be definitively excluded. Patients who have
undergone percutaneous biopsy should be considered for surgical resec-
tion if vascular invasion, atypical epithelioid component, or p53 immunore-
activity is found on histopathologic examination (Table 17.3).
Suggested Readings
Chun YS, House M, Kaur H, et al. SSAT/AHPBA joint symposium on evaluation and
treatment of benign liver lesions. J Gastrointest Surg 2013;17:636–644.
Devaney K, Goodman ZD, Ishak KG. Hepatobiliary cystadenoma and cystadenocarci-
noma. A light microscopic and immunohistochemical study of 70 patients. Am J
Surg Pathol 1994;18:1078–1091.
Ding G, Liu Y, Wu M, et al. Diagnosis and treatment of hepatic angiomyolipoma. J Surg
Oncol 2011;103:807–812.
Farges O, Ferreira N, Dokmak S, et al. Changing trends in malignant transformation of
hepatoceullar adenoma. Gut 2011;60:85–89.
Filice C, DiePerri G, Strosselli M. Outcome of hepatic amoebic abscesses managed
with three different therapeutic strategies. Dig Dis Sci 1992;37:240–247.
Hartleb M, Gutkowski K, Milkiewicz P. Nodular regenerative hyperplasia: evolving
concepts on underdiagnosed cause of portal hypertension. World J Gastroenterol
2011;17:1400–1409.
Huang C, Pitt HA, Lipsett PA, et al. Pyogenic hepatic abscess: changing trends over 42
years. Ann Surg 1996;223:600–607.
Karkar AM, Tang LH, Kashikar ND, et al. Management of hepatocellular ade-
noma: comparison of resection, embolization and observation. HPB (Oxford)
2013;15:235–243.
Lambiase RE, Deyoe L, Cronan JJ, et al. Percutaneous drainage of 335 consecu-
tive abscesses: results of primary drainage with 1 year follow up. Radiology
1992;184:167–179.
Lantinga MA, Gevers TJG, Drenth JPH. Evaluation of hepatic cystic lesions. World J
Gastroenterol 2013;19:3543–3554.
Mazza O, Fernandez DL, Pekolj J, et al. Management of nonparasitic hepatic cysts. J
Am Coll Surg 2009;209:733–739.
Mezhir JJ, Fong Y, Jacks L, et al. Current management of pyogenic liver abscess: surgery
is now second-line treatment. J Am Coll Surg 2010;210:975–983.
Russell RT, Pinson CW. Surgical management of polycystic liver disease. World
J Gastroenterol 2007;13:5052–5059.
Semelka RC, Martin DR, Balci C, et al. Focal liver lesions: comparison of dual-phase
CT and multisequence multiplanar MR imaging including dynamic gadolinium
enhancement. J Magn Reson Imaging 2001;13:397–401.
Shaked O, Siegelman E, Olthoff K, et al. Biologic and clinical features of benign solid
and cystic lesions of the liver. Clin Gastroenterol Hepatol 2011;9:547–562.
Vachha B, Sun MR, Siewert B, et al. Cystic lesions of the liver. AJR Am J Roentgenol
2011;196:W355–W366. PMID: 21427297.
Van Vledder MG, Van Aalten SM, Terkivatan T, et al. Safety and efficacy of radiofre-
quency ablation for hepatocellular adenoma. J Vasc Interv Radiol 2011;22:787–793.
WHO-Informal Working Group on Echinococcus. International classification of
ultrasound images in cystic echinococcosis for application in clinical and field
epidemiological settings. Acta Trop 2008;85:253–261.
002086384.INDD 220 7/15/2014 11:13:43 AM

18 Complications of Liver Surgery
Including Postoperative
Hepatic Insufficiency
Ching-Wei D. Tzeng and Thomas A. Aloia
INTRODUCTION
Until the most recent two decades, hepatectomies were considered high-
risk operations with frequent hemorrhagic events and high rates of morbid-
ity and mortality that would be unacceptable by current standards. Even
as surgeons have become more aggressive with the extent and radicality of
hepatectomies, the overall mortality rate has decreased dramatically from
historical rates of 10% to 20% to around 2.5%. This mortality improvement
is due to better patient risk stratification and selection, optimization of the
future liver remnant (FLR), using portal vein embolization (PVE), limiting
preoperative chemotherapy toxicity, more advanced surgical techniques,
more experienced perioperative care, novel multidisciplinary sequenc-
ing strategies, and improved capabilities to rescue patients who develop
complications.
Although the hepatectomy mortality rate has decreased dramati-
cally, the rates of any morbidity (30% to 45% by retrospective analyses) and
of major morbidity (20% by the American College of Surgeons National
Surgical Quality Improvement Program, ACS NSQIP, analysis) still remain
clinically significant. Major posthepatectomy complications include those
typical for major general surgery operations, such as bleeding, infected fluid
collections, bile leaks, wound infections, cardiopulmonary events, organ
system failure, sepsis, and venous thromboembolism (VTE). Complication
rates often correlate with the extent of hepatectomy, making reports of
overall complication rates uninformative without specification of the types
of hepatectomy examined in these studies (Table 18.1).
Three complications specific to liver surgery are bleeding (intraopera-
tive and early postoperative), bile leak (and associated organ space infec-
tion), and postoperative hepatic insufficiency (PHI, or liver failure) with
liver-related mortality. Generally, PHI can be split into “early” and “late”
phases, in which the late phase often leads to an irreversible cascade toward
liver-related mortality that can happen well past 30 postoperative days. Late
PHI mandates that surgeons look beyond the typical postoperative interval
of 30 days to completely capture all liver-related morbidity and mortality. In
fact, up to one-third of posthepatectomy deaths occur between 30 and 90
postoperative days.
Predictors of major complications can be divided into three catego-
ries—patient comorbidities, biochemical abnormalities, and perioperative
risk factors (including magnitude of operation). These factors generally
include the American Society of Anesthesiologists (ASA) class, smoking,
elevated alkaline phosphatase, low albumin, elevated partial thrombo-
plastin time (PTT), extent of hepatectomy, intraoperative or postoperative
transfusions, and prolonged operative time. The last three factors are closely
related, and thus surgeons should be cautious about pairing major simulta-
neous operations with major hepatectomies. While not all risk factors are
221
0002086385.INDD 221 7/15/2014 11:18:44 AM

TABLE
18.1 Common 30-Day Posthepatectomy Complications
All Partial Left Right Extended

Organ space infectiona 6.0% 4.5% 5.2% 7.8% 10.9%
Pneumoniaa 4.0% 3.1% 3.5% 5.5% 7.0%
Ventilator >48 ha 4.4% 3.2% 3.0% 6.7% 7.4%
Acute renal failure 1.5% 1.0% 0.4% 3.1% 2.2%
DVTa 2.0% 1.3% 1.7% 2.7% 5.2%
PEa 1.6% 1.2% 0.4% 2.5% 3.5%
Sepsisa 6.7% 5.7% 5.2% 8.4% 9.6%
Cardiac arrest 1.1% 0.7% 0.4% 2.0% 1.7%
Myocardial infarction 0.3% 0.4% 0% 0.4% 0.4%
Return to OR 4.8% 3.4% 4.8% 7.3% 7.9%
Death 2.5% 1.8% 0.9% 3.7% 5.2%
a
Complication rates increased with the extent of hepatectomy (p < 0.05)
Columns divided by magnitude of hepatectomy; DVT, deep venous thrombosis; PE, pulmonary
embolus; OR, operating room.
completely reversible, certainly many are potentially modifiable through

preoperative optimization of comorbidities and choosing operations of
lesser extent when oncologically feasible.
Besides the typically reported surgical quality outcomes of postopera-
tive complications and death, surgeons should also understand and further
study the secondary implications of major morbidity—failure to return to
preoperative performance status and failure to achieve intended oncologic
therapy. The long-term benefits of the most extraordinary and aggressive
hepatectomy may be completely negated if the patient loses his or her inde-
pendence and becomes bound to a nursing home, is unable to return to his
or her previous functional state, or is unable to complete his or her planned
adjuvant cancer therapy.
The goal of this chapter is to describe the diagnosis, management, and
risk factors for three major liver-related complications specifically associ-
ated with hepatic resection.
INTRAOPERATIVE AND POSTOPERATIVE HEMORRHAGE

Definition
Historically, hemorrhage has been the most feared complication of hepatec-
tomy. “Bleeding” broadly encompasses a range of complications including
catastrophic intraoperative hemorrhage, intraoperative bleeding requiring
intraoperative and recovery room transfusions, postoperative hemorrhage
requiring transfusions or even return to the operating room, symptomatic
anemia requiring postoperative transfusions, or small drops in hemoglobin
with no clinical sequelae.
Based on retrospective institutional reports, the posthepatectomy hem-
orrhage (PHH) rate ranges from 1% to 8%. Because of its broad definition,
the International Study Group of Liver Surgery (ISGLS) attempted to define
PHH based on transfusion requirement and clinical severity. The proposed
definition is a hemoglobin drop greater than 3 g/dL after completion of sur-
gery and/or transfusion for any postoperative hemoglobin drop and/or inva-
sive intervention (interventional radiology [IR] or surgery) to stop bleeding
(Table 18.2). Detection of hemorrhage can be via frank blood loss apparent in
perihepatic drains (drain fluid hemoglobin >3 g/dL), examination consistent
002086385.INDD 222 7/15/2014 11:18:44 AM

0002086385.INDD 223
TABLE
International Study Group on Liver Surgery Definitions of Posthepatectomy Complications
18.2 Definition Grade A Grade B Grade C
Post-hepatectomy >3 g/dL fall in hemoglobin or any ≤2 units PRBC >2 units PRBC Requires intervention
hemorrhage (PHH) postoperative transfusion for bleeding or (IR or surgery)
reintervention for bleeding
Posthepatectomy bile leak Drain bilirubin 3× serum bilirubin on or after No change from routine Requires nonoperative Requires surgical
POD3 or biloma requiring radiologic or care plan invasive intervention intervention
operative intervention or bile peritonitis
Posthepatectomy liver Impairment of liver synthetic, excretory, and No change from routine Clinical care deviates Requires intervention
failure or Postoperative detoxifying abilities on or after POD5 or care plan from routine course (IR or surgery)
hepatic insufficiency (PHI) elevation in INR or bilirubin
PRBC, packed red blood cells; IR, interventional radiology; POD, postoperative day; INR, international normalized ratio.
Chapter 18 / Complications of Liver Surgery Including Postoperative Hepatic Insufficiency 223
7/15/2014 11:18:44 AM
with intra-abdominal hematoma, or active bleeding with any radiographic

diagnosis. The ISGLS definition of PHH is specific to the postoperative period
and does not include patients who receive up to 2 units packed red blood
cells (PRBC) immediately after surgery in the recovery room. Functionally,
PHH is split into three grades, A to C. Grade A PHH is defined as transfusion
of ≤2 units PRBC with no significant change in clinical management. Grade
B PHH corresponds to transfusion of greater than 2 units PRBC, often with
changes in management including further radiographic evaluation. Bleeding
that requires any invasive intervention is categorized as grade C. Importantly,
grades B and C PHH can be associated with mortality rates as high as 17%
and 50%, respectively.
The ACS NSQIP defines major postoperative transfusion as admin-
istration of greater than 4 units PRBC after leaving the operating room.
Currently, the national rates of any intraoperative transfusion, greater than
4 units intraoperative PRBC transfusion, and major postoperative transfu-
sion greater than 4 units PRBC are 26.4%, 9.0%, and 0.8%, respectively. This
nationwide sample suggests that a significant intraoperative use of blood
products exists, but the major PHH seems to be rare.
Diagnosis and Management

The diagnosis of bleeding or hemorrhage involves clinical evaluation sup-
plemented by laboratory data, drain output (if available), and radiographic
imaging when indicated. Clinically, the patient may first develop tachycar-
dia (unless beta-blocked) and/or oliguria, with hypotension and diapho-
resis only coming later as worsening hypovolemic shock ensues. As it is
atypical for a postoperative liver surgery patient to need a large volume of
intravenous fluid, if a patient has required several fluid boluses in the early
postoperative period, a necessary workup into reasons for hypotension
or low urine output is required. While ruling out extrinsic causes (exces-
sive narcotics or epidural anesthesia) and cardiovascular events, bleeding
should be high in the differential. As bleeding becomes more likely, the
question turns to clinical significance.
The abdominal examination may not be helpful as most patients will
have incisional tenderness. The exception is with massive hemoperito-
neum that should be readily apparent and warrants urgent measures. A
change in the character of the abdominal drain fluid to a more sanguin-
eous consistency may be a sign of bleeding that can be confirmed with
serum and drain hemoglobin measurement. While the ISGLS PHH defi-
nition includes drain hemoglobin of greater than 3 g/dL, in many cases
when active bleeding is apparent, no fluid sample is needed to secure the
diagnosis. The criteria for transfusion depend on the severity of anemia
(serum hemoglobin <7 g/dL or relative drop that is hemodynamically
relevant), the degree of coagulopathy, and the patient’s overall condi-
tion. Given the evidence that transfusions may be associated with post-
operative complications, which may in turn be detrimental to oncologic
outcome, routine transfusions should be discouraged unless warranted
by hemodynamic compromise with end-organ dysfunction. A contained
hematoma does not require immediate intervention if tamponade stops
further bleeding. In a stable patient with clinical evidence of bleeding,
radiographic imaging with angiography (either by IR angiogram or by
computed tomography [CT], angiogram) may allow for diagnosis and
therapeutic intervention. However, if the patient is unstable or if IR can-
not arrest the bleeding, then a return to the operating room is necessary.
Laboratory values including platelet count, prothrombin time (PT), PTT,
and fibrinogen help guide the resuscitation and replacement of deficient
coagulation factors.
002086385.INDD 224 7/15/2014 11:18:44 AM

Preoperative and Intraoperative Considerations

Preoperative risk factors for major postoperative transfusion needs
(defined by ACS NSQIP as >4 units PRBC) include preoperative bleeding
disorder, ASA class ≥3, low albumin, and preoperative anemia. Right and
extended hepatectomies are also at greater risk for PHH. One of the key
reasons mortality from hemorrhage is now less common is better under-
standing of hepatic vascular anatomy and more advanced parenchymal
transection techniques, either with or without inflow control. The numer-
ous methods of parenchymal dissection and transection are discussed in
Chapter 13. A few salient technical issues are reviewed here. Although the
general use of stapling devices for parenchymal transection may be associ-
ated with higher bleeding rates, the targeted use of endovascular staplers to
control vascular pedicles may increase the speed and safety of liver surgery.
Newer tissue dissectors, including ultrasonic and water-jet devices, facili-
tate precise parenchymal dissection to identify vascular and biliary struc-
tures, which can be clipped or tied precisely rather than blindly transected
without ligation. Tissue-sealing devices, borrowed from minimally invasive
surgery, also aid in coagulation of the parenchymal before division.
Good planning and communication with the anesthesia team before
and during surgery facilitate the safety of hepatectomy. During the time-
out process, a discussion among the surgeon, nursing staff, and anesthe-
sia team, of the extent of hepatectomy, plan for inflow clamping (Pringle
maneuver), predicted blood loss, and blood product availability and anti-
body status, can decrease intraoperative surprises.
The concept of “low central venous pressure (CVP) anesthesia” is rec-
ommended for liver resections. This concept is critical to reducing intra-
operative hemorrhagic events. However, accurate monitoring of CVP does
not require a central venous catheter. Regardless of the method of intra-
vascular volume monitoring, intravenous fluid administration should
remain low (the surgeon can aid in monitoring this by examining the infe-
rior vena cava) until the transection is completed. Keeping the CVP low
significantly reduces venous bleeding during the parenchymal transection.
Laparoscopic liver surgery has been associated with decreased blood loss
because the required pneumoperitoneum decreases venous oozing from
the parenchymal surface. The use of stay stitches to elevate the liver out
of the abdominal cavity (with laparotomy pads behind the mobilized liver)
may elongate and compress hepatic veins, decreasing venous bleeding dur-
ing transection.
Regardless of the parenchymal dissection and transection technique,
after the resection, the liver edge should be carefully inspected for vessels or
bile ducts that were not adequately ligated. A gentle Valsalva maneuver by
the anesthesia team can expose hepatic vein tributary bleeding, which may
require suture ligation. The role of topical hemostasis to the liver transection
bed with cautery, coagulant sheets, or fibrin glue is frequently debated. With
no overt bleeding sites, some surgeons leave the surface alone, arguing that
a viable liver edge may help decrease postoperative bile leak. Others cauter-
ize the entire cut surface with or without placement of a hemostatic agent
( fibrin glue or cellulose netting/gauze) applied to the surface. Others take
the time to develop and place an omental flap to cover the raw liver surface.
Although the elements that contribute to major bleeding events in
liver surgery are complex, an operative strategy that identifies and controls
major hepatic vein tributaries and portal triad branches before and during
the parenchymal dissection is the best way to avert major bleeding (and
not coincidentally bile leak as well). Such a strategy is facilitated by care-
ful preoperative planning that includes a thorough understanding of the
patient’s liver anatomy including the relationship of the liver vasculature
0002086385.INDD 225 7/15/2014 11:18:44 AM

to the target lesion(s). High-quality three-phase liver protocol CT scan or

high-resolution magnetic resonance imaging (MRI) is important. General
abdominal CT scans (such as those ordered during emergency room evalu-
ation for abdominal pain) are inadequate and may be dangerous if used as
a guide. Intraoperative ultrasound is also essential to permit exact dissec-
tion along the correct planes. This precise dissection balances adequate
oncologic margins with maximal parenchymal sparing when possible,
dividing only the necessary vessels and biliary pedicles. While bleeding risk
cannot be fully eliminated, with adequate understanding of the patient’s
liver anatomy, good communication with the entire operating room team,
and proper use of available intraoperative surgical technology, the risk of
intraoperative and postoperative hemorrhage can be purposefully reduced
to minimal rates.
BILE LEAK AND ORGAN SPACE INFECTIONS

Definition
Despite an overall decrease in posthepatectomy complications during
the past two decades, the problem of bile leak still remains unsolved. In
modern series, the incidence of postoperative bile leak ranges from 2% to
33%, depending on the reported series, definition, and extent of hepatec-
tomy. Like high-grade PHH, high-grade bile leak remains a major cause
of associated morbidity, often leading to longer hospitalization, increased
health care costs, prolonged disability with abdominal drains, and the need
for additional procedures. The mortality associated with bile leak can be
as high as 39%. Data on rates of bile leak are varied because like PHH, a
variety of bile leak definitions have been used. Some perihepatic fluid col-
lections remain undrained and are treated with antibiotics only, further
limiting the ability to accurately label them as “bilomas” in retrospective
series. However, after liver resection, most organ space infections can be
presumed to be bilomas or bile leaks.
The ISGLS convened an expert panel to define three grades of bile
leak after hepatectomy. Mirroring the trend to grading surgical complica-
tions based on clinical sequelae rather than arbitrary laboratory values of
bilirubin or international normalized ratio (INR), the ISGLS divided bile
leaks into three categories. Bile leak was globally defined as a drain biliru-
bin at least three times the serum bilirubin on or after postoperative day
3, a biloma requiring radiologic or operative intervention, or clinically evi-
dent bile peritonitis. Grade A bile leak does not affect clinical management.
Grade B requires a deviation in typical postoperative management such as
a percutaneous drainage, but does not require surgery. Grade C bile leaks
require a return to the operating room and are associated with extremely
high mortality rates.

The management of a bile leak can be summarized by the singular impor-
tance of timely diagnosis and drainage if needed. Delays in diagnosis and
drainage only feed the inflammatory cascade by allowing sepsis to trigger a
systemic inflammatory response syndrome. In general, with proper external
drainage and control of infection with antibiotics, bile leaks will heal with
time as the surgical bed recovers from operative trauma. When left unat-
tended and discovered late, bile leaks can be highly morbid requiring longer
hospitalizations, more interventions, and transfers to higher-acuity beds.
Bile leaks with high mortality rates are typically from associated complica-
tions. As an example of its significant downstream effects, organ space infec-
tion is the risk factor most strongly associated with posthepatectomy VTE.
002086385.INDD 226 7/15/2014 11:18:44 AM

Bile leak, also known as biloma or biliary fistula, can be diagnosed

by several methods depending on the timing, severity, and presence or
absence of an intraoperatively placed surgical drain. In the presence of a
drain, the character of the fluid, which, if necessary, can be tested for bili-
rubin, can readily diagnose bile leak. By the ISGLS definition, a fluid bili-
rubin three times the serum bilirubin constitutes a bile leak. If the drain
works well and there is no evidence of sepsis from an undrained biloma,
then no further imaging is necessary. The drain is simply left in place when
the patient is discharged from the hospital until the fluid output drops to a
minimum. Antibiotics are not necessary in the absence of sepsis, abdomi-
nal wall infection, or undrained organ space infection. If the drain output
does not decrease with short-term outpatient management, then either an
endoscopic retrograde cholangiopancreatogram (ERCP) or percutaneous
transhepatic cholangiogram (PTC) is warranted to study where the leak is
coming from (central vs. peripheral bile ducts) and to place a biliary stent
to increase preferential bile drainage toward its natural outflow tract (duo-
denum or enteric anastomosis). Neither ERCP nor PTC needs be reflexively
ordered when bile leak is first diagnosed and drain output has not yet been
defined, because both ERCP and PTC have their associated risks.
If no surgical drain is present or if the bile leak is diagnosed after the
drain has been removed, then CT is warranted to identify the location of the
fluid collection. Because these bile leaks have been undrained for several
days, the patients will often present with pain, fever, leukocytosis, nausea,
ileus, and/or abdominal discomfort. Associated right pleural effusions may
also occur. With systemic symptoms and leukocytosis, antibiotics are usually
necessary. Antibiotics may be quickly tailored based on culture data. The CT
allows the surgeon to determine if IR can access the presumed biloma and
place a percutaneous drain. CT is important because it can help differenti-
ate infected fluid collections with a rind or air bubbles versus benign fluid
collections that are likely seromas not requiring drainage. In some cases,
liver compression by a perihepatic fluid collection can be appreciated and
helps to identify an ongoing bile leak “under pressure” (Fig. 18.1). Because of
their later diagnosis, these patients usually require a short inpatient admis-
sion for CT, percutaneous drainage, and antibiotics. After the sepsis is con-
trolled, they can be discharged and managed in a similar fashion as patients
who had a bile leak diagnosed by a surgically placed drain. Treatment of bile
leak includes timely diagnosis, control of sepsis through adequate drainage,
and confirmation (clinically or radiologically) of preferential anatomic flow
of bile toward the duodenum or through an enteric anastomosis.

Perioperative risk factors for bile leak include bile duct resection/recon-
struction, extended hepatectomy, repeat hepatectomy, need for en bloc
diaphragm resection, and intraoperative transfusion. The surgeon should
utilize a preoperative high-quality CT or MRI to define the planned extent
of resection. Part of the resection plan should include the strategy for
extrahepatic versus intrahepatic bile duct division. Using hilar cholangio-
carcinoma as an example, the extent of biliary resection, the level of recon-
struction, and the type of anastomosis should be anticipated. Preoperative
planning can help prevent both unplanned division of the contralateral
hepatic duct and accidental injury of hepatic duct tributaries, which can
lead to catastrophic postoperative bile leaks.
After the resection, the transected liver surface and the hepatic duct
division line are visually inspected for any bile staining. The liver itself
should also be inspected for areas where the capsule has been sheared off
or for areas with lacerations from retraction injury, as these can be the sites
0002086385.INDD 227 7/15/2014 11:18:44 AM

FIGURE 18.1 Posthepatectomy biloma. Liver compression by a posthepatectomy fluid col-

lection can be appreciated and helps to identify an ongoing bile leak “under pressure.”
of subsequent bile leaks. A number of posttransection cholangiogram tech-

niques have been described in the literature. These methods include intra-
ductal injections of fluoroscopic contrast (conventional cholangiography),
saline, methylene blue, “white” fat emulsion, indocyanine green (ICG), or
air (at our institution) for the intraoperative detection and repair of poten-
tial postoperative bile leaks. As with PHH prevention, some surgeons use
various forms of hemostatic agents to “seal” the transection surface, and
omental flaps can be used, although little evidence exists to support either
practice in the prevention of bile leaks. There is also no consensus on the
role of surgical drain placement. Some surgeons routinely place drains,
even for partial hepatectomies. Others reserve drains for higher-risk opera-
tions such as major/extended hepatectomies or those with biliary–enteric
anastomoses, while others do not routinely drain any hepatectomies.
When drains are placed, they should be removed at the earliest possible
time when they no longer function as a reliable indicator of bleeding or bile
leak (usually 2 to 4 days). In summary, because bile leak remains a major
source of potentially preventable morbidity, mortality, cost, and disability,
surgeons should strive to develop a systematic preventative approach to the
intraoperative identification of bile leaks and to be vigilant about the early
detection of bile leaks in the postoperative setting.
POSTOPERATIVE HEPATIC INSUFFICIENCY AND

LIVER-RELATED MORTALITY
Definition
After liver resection, the least reversible complication in the recovery
period is PHI, or posthepatectomy liver failure. Its clinical consequences
range from mild biochemical derangements with no change in routine
postoperative course to irreversible liver failure with death. Ultimately,
PHI is related to inability of the FLR to perform the functions of protein
002086385.INDD 228 7/15/2014 11:18:45 AM

synthesis, biliary drainage, and filtration. But, PHI can be divided into early
and late postoperative manifestations, each with its own risk factors. Early
PHI is from insufficient FLR volume or acute liver injury that prevents the
FLR from tolerating the immediate surgical and postoperative stress. This
injury can include intraoperative liver ischemia from long inflow occlusion
intervals without adequate reperfusion, perioperative major blood loss,
and/or perioperative systemic hypotension, which combined with mar-
ginal FLR volume can lead to early liver failure. Beyond the acute phase of
recovery, late failure is associated with insufficient FLR regeneration. Most
commonly, this phenomenon is related to intrinsic parenchymal injury
from preoperatively underrecognized cirrhosis, steatohepatitis, and/or
extended-duration chemotherapy.
In general, PHI is defined by the ISGLS as the impairment of the liv-
er’s synthetic (INR), excretory (bilirubin), and detoxifying functions. The
ISGLS avoided the use of arbitrary serum laboratory values in their defini-
tion, focusing instead on three grades of clinical sequelae. Grade A causes
no change in routine postoperative care. These patients might have mild
cholestasis, which does not affect their activities of daily living or their dis-
charge planning. Grade B PHI requires some deviation from the routine
clinical course. Grade C PHI requires invasive intervention.
A more clinically relevant definition of PHI was proposed in a retro-
spective international study of over 1,000 patients from three institutions
undergoing hepatectomy of ≥3 segments. PHI was defined as a peak
total bilirubin greater than 7 mg/dL, which had a sensitivity and speci-
ficity of greater than 93% and an odds ratio of 10.8 for predicting 90-day
mortality. The authors found this cutoff to be a better predictor of liver-
related mortality than previous combinations of hyperbilirubinemia
with elevated PT.

There are a limited number of effective therapies to rescue patients from
severe liver failure after hepatectomy. If PHI is suspected, the management
includes ruling out biliary obstruction and biloma as reasons for increased
serum bilirubin. If the serum bilirubin remains high despite a patent bile
duct or biliary anastomosis, then cholestasis from PHI is the diagnosis. In
this state, patients are especially tenuous and unlikely to tolerate further
physiologic insults such as sepsis. As liver regeneration is dependent on
nutrition and a lack of sepsis, nutritional support and timely treatment of
any infections are especially important for recovery. Burdened by ascites,
many patients will need diuresis to control their abdominal symptoms and
to facilitate their ability to participate in activities of daily living. If there
is no sepsis and no further need for inpatient management, these patients
can be followed as outpatients. Labs every few days to weekly are sufficient
to measure the trends, but it may take weeks to months for the late PHI to
resolve. If the liver function tests and bilirubin do not eventually plateau,
then death from liver failure is inevitable.

Given the limited ability to reverse PHI, the best management of this com-
plication is prevention. To avoid this complication, the surgeon must focus
on preoperative planning and intraoperative considerations in choosing
the extent of hepatectomy based on the FLR size and functional capac-
ity. The FLR is the predicted volume of liver after resection, and adequate
size, synthetic function, biliary drainage, and regenerative capacity are
the keys to avoiding both early and late PHI. Based on a patient’s calcu-
lated total liver volume (TLV) standardized to body surface area (BSA)
0002086385.INDD 229 7/15/2014 11:18:45 AM

(TLV cm3 = −794.41 + 1,267.28 × BSA m2), the predicted FLR is compared as

a ratio to this standardized TLV calculation to yield the standardized FLR.
Generally, the required minimum standardized FLR is greater than 20% for
normal livers, greater than 30% for livers with limited parenchymal injury
such as from chemotherapy, and greater than 40% for permanently dam-
aged livers such as those with cirrhosis. A number of modifiable risk factors
may decrease PHI from inadequate FLR.
Arguably, the greatest advance in preventing early and late PHI in the
past decade has been greater utilization of PVE before a major/extended
hepatectomy. With increased availability of CT volumetry, the FLR volume
can be accurately calculated down to the decimal point, and “eyeballing”
a CT scan to estimate FLR is now discouraged. Up to 10% of patients at
initial presentation have insufficient FLR for a right hepatectomy, and as
many as 75% will have an inadequate FLR for an extended right hepatec-
tomy. In practice, the real number is higher for both, because these calcu-
lations were based on patients with normal liver parenchyma and normal
FLR volume requirements.
One may think of PVE as a functional test of the FLR that allows
assessment of regenerative capacity prior to committing to hepatic resec-
tion. During the time between PVE and hepatectomy, surgeons can study
both the degree of hypertrophy (DH, needs to be >5% points) and the kinetic
growth rate (KGR = DH divided by time in weeks) of the FLR. Importantly,
the technical aspects of PVE should be carefully considered to achieve max-
imum hypertrophy. With complete PVE of the right portal vein and segment
IV branches with spherical microparticles and coils, an FLR hypertrophy of
up to 69% can be expected.
In some cases with bilateral and/or multifocal disease, safe hepatec-
tomy cannot be accomplished with a single operation. In this setting, the
two-stage approach, with first-stage partial left hepatectomy, followed by
percutaneous PVE, and second-stage extended right hepatectomy, has
achieved excellent oncologic outcomes. As it facilitates extensive liver
resection while limiting PHI, it is important to add the two-stage approach
to the surgeon’s armamentarium.
A major modifiable risk factor for late PHI is the duration of preop-
erative chemotherapy given to patients with liver tumor. The goal of pre-
operative chemotherapy should not be “optimal response” with complete
disappearance of the liver lesions because extended-duration chemo-
therapy (>8 cycles) can increase the risk of chemotherapy-associated liver
injury (CALI) and postoperative complications. Steatohepatitis from irino-
tecan (usually in patients with metabolic syndrome phenotype) is particu-
larly important as it is the only CALI associated with increased mortality
from PHI. This may be a key consideration when choosing neoadjuvant
treatment with between FOLFOX and FOLFIRI. All of the aforementioned
modifiable factors require preoperative planning by the surgeon in the con-
text of a multidisciplinary discussion with medical oncologists, patholo-
gists, hepatologists, and radiologists.
OUTCOMES/FOLLOW-UP
Posthepatectomy complications have significant short-term and long-
term consequences. Certainly, the most impactful short-term complica-
tion is death, but fortunately, mortality from hepatectomy is infrequent.
Early mortality from PHH is rare because of improved knowledge of liver
anatomy, better preoperative imaging, intraoperative ultrasound, and bet-
ter parenchymal dissection/transection tools. Late PHI-related mortality
is better understood, and the necessary preoperative preventative steps
002086385.INDD 230 7/15/2014 11:18:45 AM

(patient selection, CT volumetry ± PVE, and limiting chemotherapy) are

being used more often.
As hepatectomy becomes more routine, outcome studies should move
beyond 30-day metrics to include collection of data on return to intended
oncologic therapy and return to preoperative functional status. Both mea-
sures more completely describe the downstream effects of complications
beyond the mortality rate, and both measures may account for some of
the data that link transfusions and major complications with decreased
survival in cancer patients undergoing hepatectomy. Improvement in the
diagnostic methods and the management of posthepatectomy complica-
tions have dramatically increased the rescue rate of patients from mortality
after complicated hepatectomy. The next step in progress is to further opti-
mize preoperative and intraoperative measures to prevent postoperative
complications.
Suggested Readings
Abdalla EK, Denys A, Chevalier P, et al. Total and segmental liver volume variations:
implications for liver surgery. Surgery 2004;135(4):404–410.
Abdalla EK, Hicks ME, Vauthey JN. Portal vein embolization: rationale, technique and
future prospects. Br J Surg 2001;88(2):165–175.
Aloia TA, Fahy BN, Fischer CP, et al. Predicting poor outcome following hepatec-
tomy: analysis of 2313 hepatectomies in the NSQIP database. HPB (Oxford)
2009;11(6):510–515.
Aloia TA, Zorzi D, Abdalla EK, et al. Two-surgeon technique for hepatic parenchymal
transection of the noncirrhotic liver using saline-linked cautery and ultrasonic
dissection. Ann Surg 2005;242(2):172–177.
Brouquet A, Abdalla EK, Kopetz S, et al. High survival rate after two-stage resection
of advanced colorectal liver metastases: response-based selection and complete
resection define outcome. J Clin Oncol 2011;29(8):1083–1090.
Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new
proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann
Surg 2004;240(2):205–213.
Kishi Y, Abdalla EK, Chun YS, et al. Three hundred and one consecutive extended right
hepatectomies: evaluation of outcome based on systematic liver volumetry. Ann
Surg 2009;250(4):540–548.
Kishi Y, Zorzi D, Contreras CM, et al. Extended preoperative chemotherapy does
not improve pathologic response and increases postoperative liver insuffi-
ciency after hepatic resection for colorectal liver metastases. Ann Surg Oncol
2010;17(11):2870–2876.
Koch M, Garden OJ, Padbury R, et al. Bile leakage after hepatobiliary and pancreatic
surgery: a definition and grading of severity by the International Study Group of
Liver Surgery. Surgery 2011;149(5):680–688.
Kooby DA, Stockman J, Ben-Porat L, et al. Influence of transfusions on perioperative
and long-term outcome in patients following hepatic resection for colorectal
metastases. Ann Surg 2003;237(6):860–869; discussion 869–870.
Madoff DC, Abdalla EK, Gupta S, et al. Transhepatic ipsilateral right portal vein embo-
lization extended to segment IV: improving hypertrophy and resection outcomes
with spherical particles and coils. J Vasc Interv Radiol 2005;16(2 Pt 1):215–225.
Mullen JT, Ribero D, Reddy SK, et al. Hepatic insufficiency and mortality in 1,059 non-
cirrhotic patients undergoing major hepatectomy. J Am Coll Surg 2007;204(5):
854–862; discussion 862–854.
Rahbari NN, Garden OJ, Padbury R, et al. Post-hepatectomy haemorrhage: a defini-
tion and grading by the International Study Group of Liver Surgery (ISGLS). HPB
(Oxford) 2011;13(8):528–535.
Rahbari NN, Garden OJ, Padbury R, et al. Posthepatectomy liver failure: a definition
and grading by the International Study Group of Liver Surgery (ISGLS). Surgery
2011;149(5):713–724.
Ribero D, Abdalla EK, Madoff DC, et al. Portal vein embolization before major hepa-
tectomy and its effects on regeneration, resectability and outcome. Br J Surg
2007;94(11):1386–1394.
0002086385.INDD 231 7/15/2014 11:18:45 AM

Tzeng CW, Aloia TA. Colorectal liver metastases. J Gastrointest Surg 2013;17(1):
195–201.
Tzeng CW, Katz MH, Fleming JB, et al. Risk of venous thromboembolism outweighs
post-hepatectomy bleeding complications: analysis of 5651 National Surgical
Quality Improvement Program patients. HPB (Oxford) 2012;14(8):506–513.
Zimmitti G, Roses RE, Andreou A, et al. Greater complexity of liver surgery is not
associated with an increased incidence of liver-related complications except for
bile leak: an experience with 2,628 consecutive resections. J Gastrointest Surg
2013;17(1):57–64.
002086385.INDD 232 7/15/2014 11:18:45 AM

19 Multidisciplinary Approach
to Liver Oncology
Mary A. Maluccio
MULTIDISCIPLINARY APPROACH TO LIVER ONCOLOGY

Liver oncology has expanded over the past decade to include not only the
primary liver malignancies for which treatment paradigms have been estab-
lished but also the secondary malignancies for which treatment paradigms
are in evolution. Unlike other multidisciplinary oncology programs that
have three major physician-based pillars, medical oncology, surgical oncol-
ogy, and radiation oncology, adequate care of the liver oncology patient
requires considerably broader physician engagement including input and
commitment from organ transplant, hepatology, interventional radiology,
abdominal imaging, and nuclear medicine. As we move into a health care
environment that will expect well-defined treatment algorithms based on
best practices, practitioners will need to move away from institutional-
specific paradigms and embrace a more uniform approach that takes into
consideration both effectiveness of treatment, quality outcomes, utilization
of health care resources, and cost. An understanding of the advantages and
disadvantages of treatment platforms and how patient- and tumor-specific
variables weigh into medical decision making is the objective of this chapter.
PRIMARY LIVER CANCER (HEPATOCELLULAR CARCINOMA AND

CHOLANGIOCARCINOMA)
Primary liver cancer presents unique challenges for which a multidisci-
plinary approach to patient care is considered imperative and without
which quality outcome parameters will be difficult to reach. Hepatocellular
carcinoma (HCC) is one of the few cancers with increasing incidence. As
such, there has been a focus on safe and accurate diagnosis and the devel-
opment of treatment algorithms that take into consideration the unique
complexities of this patient population. The past decade has seen improve-
ments in nonsurgical treatment platforms and better standardization with
respect to the diagnosis and eligibility for liver transplant. How to navigate
patients through the challenges of treatment is difficult and depends on
several factors: (1) patient-related variables such as comorbid conditions
that influence treatment eligibility; (2) liver-related variables such as Child-
Pugh score; and (3) tumor-related variables such as size, number, pattern of
spread within the liver, and vascular involvement.
Five-year survival rates for HCC in the United States have improved
modestly to approximately 26%; felt to be associated with improved surveil-
lance in identifiable high-risk patients (i.e., hepatitis B and C) and surgical
intervention (i.e., resection or transplant) for early-stage disease. The vast
majority of HCC occurs in the setting of chronic liver disease from viral hepa-
titis, alcohol abuse, and/or nonalcoholic steatohepatitis (NASH). Prevention
of HCC must therefore focus on the prevention of hepatitis B and C trans-
mission and the institution of guidelines to reduce the prevalence of obesity.
233
0002100649.INDD 233 7/14/2014 7:19:28 PM

Consensus guidelines have been published by several o rganizations

including the American Association for the Study of Liver Disease
(AASLD), National Comprehensive Cancer Network (NCCN), and European
Association for Study of Liver (EASL) to standardize the approach to diagno-
sis and treatment. As is true with the most disease processes, HCC is more
effectively treated when it is diagnosed at an early stage. The best chance
for prevention and/or early diagnosis comes from attempts to eliminate
viral hepatitis and the surveillance of patients known to be at high risk. This
includes patients with cirrhosis from any cause and hepatitis B carriers. The
2012 NCCN guidelines recommend screening high-risk patients with serum
AFP and liver ultrasound every 6 to 12 months. A rising AFP associated with
a liver nodule greater than 1 cm raises suspicion for HCC and warrants eval-
uation with cross-sectional imaging. Better biomarkers of cancer risk would
improve our ability to stratify patients into more cost-effective surveillance
programs, chemoprevention trials, and/or treatment algorithms.
Criteria for the diagnosis of HCC have evolved over the past decade.
In order to minimize the use of percutaneous biopsy and its inherent risks
in patients with underlying liver disease (tract seeding, bleeding, etc.), the
AASLD, NCCN, and EASL working groups have adopted imaging criteria
that predict cancer with acceptable accuracy. Dedicated abdominal imag-
ers are necessary to protocol and read the cross sectional imaging such that
the reports are in keeping with what is required to appropriately character-
ize the lesions. On contrast-enhanced images using CT and MRI, the typical
enhancement pattern of HCC is early arterial enhancement and venous phase
washout, related to the fact that these are hypervascular lesions supplied
predominantly by branches of the hepatic artery. In the setting of chronic
liver disease, lesions larger than 1 cm that demonstrate these imaging char-
acteristics on triple-phase CT or contrast-enhanced MRI are considered to be
HCC. This is a change from previous guidelines where lesions between 1 and
2 cm required characteristic enhancement on both imaging modalities (CT
and MRI) to define HCC. Despite changes in the imaging criteria, only lesions
greater than 2 cm with characteristic enhancement are eligible for model of
end-stage liver disease (MELD) exception points for liver transplant. Some
centers have adopted MRI with a novel contrast agent, gadoxetate disodium
(EOVIST), to better define lesions that do not meet criteria on arterial and
venous-phase imaging alone. Lesions suspicious for HCC appear darker
than background liver on T1- (hepatocyte-phase) weighted imaging. To date,
EOVIST has not changed the diagnostic paradigm currently used to deter-
mine treatment eligibility despite reports of improved imaging specificity.
Several clinical staging systems, including the Cancer of the Liver
Italian Program (CLIP) and the Barcelona Clinic Liver Cancer (BCLC), have
emerged to predict prognosis and stratify patients for treatment. The goal
of each staging system is the same: to better define the prognostic weight
of clinical variables on outcome in HCC patients being considered for treat-
ment or clinical trials. The CLIP system includes Child-Pugh score, tumor
morphology (uninodular, multinodular, or extensive), AFP, and the pres-
ence or absence of portal vein thrombosis. The BCLC system includes the
Child-Pugh score, clinical performance status, and tumor stage (solitary,
multinodular, vascular invasion, or extrahepatic spread) and categorizes
patients. Early HCC (BCLC A1-A4) includes well-compensated (Child A)
liver reserve with an excellent performance status and limited tumor
burden. Intermediate HCC (BCLC B) includes moderate liver reserve
(Child A and B), excellent performance status, and multinodular tumors.
Advanced HCC (BCLC C) includes patients with moderate liver reserve
(Child A and B), v ascular invasion or extrahepatic spread, and a vulnerable
performance status (PST 1–2). The difference in estimated survival at
002100649.INDD 234 7/14/2014 7:19:28 PM

Chapter 19 / Multidisciplinary Approach to Liver Oncology 235
TABLE
19.1 Treatment Options for HCC
Liver transplant
Liver resection
Ablation
Thermal (radiofrequency, microwave, etc.)
Chemical (ethanol, acetic acid)
Transarterial therapy
Bland embolization
Chemoembolization
Radioembolization
SBRT
Chemotherapy
3 years in untreated BCLC A versus C patients was 50% versus 8%,

respectively. The impact of any given treatment on patients with more
advanced disease is unclear. Modifications of these staging systems with
the addition of plasma-based tumor markers like vascular endothe-
lial growth factor or insulin-like growth factor-1 have been proposed to
improve prognostic stratification of patients with advanced HCC and better
select patients for treatment or clinical trials.
Multiple treatment options are available for patients with HCC
(Table 19.1).
Liver Transplant
Liver transplant is considered the most effective method to treat both the
cancer and the underlying liver disease from which most HCC develop.
Therefore a productive relationship between oncology (surgical and medi-
cal) and transplant is required to maximize this option for appropriate
patients. Transplant eligibility is based on the size and number of tumors,
and criteria have been established to optimize cancer-specific outcomes.
Most commonly used worldwide are the Milan criteria in which patients
with up to three foci of HCC less than 3 cm or one tumor less than 5 cm
are eligible for liver transplantation. These patients experienced 5-year
overall survival (75%) that paralleled survival observed in patients trans-
planted without cancer at that time. Other centers, such as The University
of California San Francisco (UCSF), have broadened criteria (one tumor
<6.5 cm or two to three tumors, none >4.5 cm, with the total tumor diameter
not to exceed 8 cm) for eligibility based on outcome-based evidence that
less strict parameters do not adversely affect overall survival with the devel-
opment of more sophisticated liver-directed therapies (LDTs) for HCC, and
down-staging of patients into either Milan or UCSF criteria has emerged
as a reasonable approach to select patients. What has become apparent is
that progression of disease despite LDTs identifies cancers that are at high
risk for recurrence after transplant. Demonstration of a response to LDTs
prior to transplant in combination with surveillance over a period of time
prior to committing to transplant allows centers to select out more favor-
able biology and broaden patient eligibility without compromising cancer-
specific survival.
Surgical Resection
Liver resection remains the gold standard for patients with resectable HCC
that develops in the setting of normal liver. However, most patients with
HCC have diseased liver parenchyma, and resection in this population is
0002100649.INDD 235 7/14/2014 7:19:28 PM

more fraught with potential for complications. For this reason, preservation
of liver parenchyma is critical, and treatment requires a balance between
the effect of any surgical intervention short of transplant and the poten-
tially d
etrimental effect of this treatment on a vulnerable and “high-risk”
remnant. Most published resection series focus on patients with single
tumors or limited disease burden up to a certain size and well-preserved
(Child A) function. As LDTs have improved, the gap in overall survival
between LDT and resection in patients with underlying liver disease has
narrowed substantially. This is in part due to the high rate of recurrence or
de novo tumor emergence in the liver remnant. The recurrence rate after
resection is approximately 50% at 2 years and 75% at 5 years in most series.
Institutional treatment paradigms with respect to resection versus trans-
plant should be established with input from both transplant and surgical
oncology since transplant eligibility may come into play with either recur-
rence after resection or in the event that the patient struggles with liver
insufficiency postoperatively.
In regions of the world where hepatitis B is the dominant risk fac-
tor for cancer, resection is employed more commonly for several reasons
including the following: (1) cadaveric organ availability is limited; (2) cen-
ters outside the United States rely more on living related donor pools where
the human investment in the process is greater; and (3) a higher proportion
of patients with hepatitis B have preserved liver function making resection
safer. Therefore, to minimize unnecessary risk to a living donor and help
select patients whom would most benefit from a liver transplant, resection
is used up front with transplant reserved as a salvage option in the event
that the cancer recurs or the liver function worsens over time.
The natural history of HCC in the background of NASH would suggest
a higher proportion of noncirrhotic patients and a lower rate of recurrence
(or de novo tumor emergence) than either hepatitis B or hepatitis C. For
this reason, resection may emerge as a reasonable option in this patient
population as well. Resection versus transplant in NASH patients must be
evaluated based on underlying liver reserve.
Embolization
Most patients are not candidates for resection or transplantation at the
time of diagnosis because of either the extent or the distribution of tumor,
underlying liver function, or medical comorbidities. Interventional radiol-
ogy has played a central role in the treatment of HCC for decades. For this
reason, engagement of interventional radiology is critical to adequately
address the range of HCC patients within a robust liver oncology program.
The dual blood supply to the liver has allowed the development of hepatic
artery–based therapies over the past 30 years. Whereas non–tumor-bear-
ing liver parenchyma receives nutrient supply predominantly from the
portal vein, most HCC are supplied predominantly by the hepatic artery.
Catheter-based techniques take advantage of this unusual architecture to
deliver intra-arterial therapy directly to tumor bed. Several different treat-
ments have been administered by catheter via the artery to treat HCC,
including bland embolization, transarterial chemoembolization (TACE),
chemoembolization with drug-eluting beads (DEB), and radioemboliza-
tion. To date, there have been no prospective or randomized trials defining
any of the available options as superior in terms of survival. Centers around
the world have therefore gravitated toward the technique that works best
in their hands. Complications common to all catheter-based therapies for
HCC include postembolization syndrome ( fever, nausea, and pain), non-
target embolization (stomach, gallbladder, duodenum, pancreas), and liver
failure (<2% in well-selected patients).
002100649.INDD 236 7/14/2014 7:19:28 PM

Bland Particle Embolization

Bland particle embolization is based on the unique dependence of HCC
on the hepatic artery. Small particles (40 to 120 μm) are injected into the
tumor arterial supply to cause terminal vessel blockade and resultant isch-
emic necrosis. The 5-year data on response and survival suggest compa-
rable cancer-specific outcomes to other catheter-based techniques. The
results of bland embolization are essentially immediate, and radiologic
evidence of tumor necrosis is seen within hours of the procedure. This is a
particularly useful feature in patients who present with significant tumor
burden where further progression may render them untreatable. Other
theoretical advantages of bland embolization include (1) the particles
come in a range of sizes that can effectively address unique vascular char-
acteristics of the tumor, including intrahepatic portal–systemic shunting;
(2) lower periprocedural cost; (3) no delay between initial arteriogram and
treatment delivery; (4) no chemotherapy or radiation-related side effects;
(5) the ability to retreat due to better preservation of intrahepatic arteries
after treatment; and (6) less institutional infrastructure requirements such
as radiation safety.
Chemoembolization/Drug-Eluting Beads
Doxorubicin-eluting beads are another catheter-based LDT. The use of an
eluting bead is considered an improvement on conventional chemoembo-
lization (TACE) in which hydrophilic chemotherapeutic agent(s) (with or
without Lipiodol) was injected into the liver via the hepatic artery. To pre-
vent washout of the chemotherapy from the tumor bed and thereby allow
prolonged contact between chemotherapeutic agent(s) and tumor cells, the
feeding artery was then occluded with particles or Gelfoam. Conventional
TACE has largely been replaced by embolization with DEBs. DEBs are pre-
formed deformable microspheres that are loaded with doxorubicin up
to 150 mg per treatment. The pharmacokinetic profile of the DEB is sig-
nificantly different from that seen with conventional TACE, with evidence
that the peak drug concentration in the serum is an order of magnitude
lower for DEBs compared to TACE. Objective response by EASL criteria has
been reported to be 70% to 80%. One- and 3-year survival rates of 89.9%
and 66.3%, respectively, have been reported in a heterogeneous cohort of
Barcelona Clinic Liver Cancer (BCLC) A–C patients. The advantages of DEB
overlap with those related to bland embolization: (1) the ability to treat
multiple tumors in different regions of the liver during the same procedure;
(2) the use of superselective techniques limits toxicity to normal liver sub-
stance; and (3) the ability to repeat the procedure several times over the
lifetime of the patient.
Radioembolization
Yttrium-90 (Y90) is a beta-emitter that can be loaded into glass or resin
microspheres and administered via a microcatheter in the hepatic artery.
™
The TheraSphere glass microspheres are approved by the U.S. Food and
Drug Administration (FDA) for treating HCC. The spheres are preferentially
taken up by tumor vasculature and, as such, deliver a high dose of radiation
directly to the tumor bed. The half-life of the bead allows for treatment over
weeks with the theoretical advantage of an improvement in durability of
response. Other advantages of Y90 include (1) better tolerability in patients
with vulnerable liver reserve (Child B) when used in a selective manner; (2)
because of the size and number of particles, there is little embolic effect;
(3) the effect of radiation is less acute than any of the embolic techniques,
and there is less postembolization syndromes commonly seen after TACE,
DEBs, or bland particle embolization. Y90 is d elivered in the outpatient
0002100649.INDD 237 7/14/2014 7:19:28 PM

setting. The objective response rate is c omparable to other catheter-based

modalities and depends on several factors including size of the lesion, pat-
tern of spread within the liver (unilobar vs. bilobar), and vascularity of the
lesion noted on planning arteriogram. In patients with limited liver reserve
(Child C), some centers will still consider treatment but usually as a bridge to
a timely transplant. There are disadvantages specific to radioembolization:
(1) the need for a mapping procedure to embolize potential nontarget vas-
culature arising from or near the target vessels (i.e., right gastric artery,
falciform artery, gastroduodenal artery); (2) the risk of shunting radioactive
particles into the lung resulting in pulmonary fibrosis; and (3) radiation-
induced liver toxicity (approximately 1% to 3%). Nontarget radioemboli-
zation to the lung or GI tract can be particularly devastating compared to
TACE or bland embolization.
Tumor evaluation after any catheter-based treatment is difficult.
Instead of decrease in the size and volume of tumor as seen after response to
chemotherapy, a change in enhancement from early arterial to no enhance-
ment is widely accepted as response. Therefore, the modified RECIST
criteria (mRECIST) have been established for this purpose. mRECIST
reports the percentage of tumor that has imaging findings consistent with
necrosis rather than absolute size measurements. After radioembolization,
1-month response rates are difficult to measure because there are often
radiation-induced changes in the treated liver. In cases in which AFP is
elevated, changes in AFP may provide better insight into early response.
A 3-month time point is a far better judge of maximal response to treat-
ment. Continued response to Y90 even after 3 months has also been
observed.
Stereotactic Body Radiation Therapy

Stereotactic body radiation therapy (SBRT) is a type of targeted radiation
therapy whereby computer modeling is used to delineate the treatment
area. Historically radiation therapy for HCC has been limited to the pal-
liation of pain in surgically unresectable disease. Based on contemporary
data, SBRT now enters the most recent NCCN guidelines as a reasonable
up front treatment in select patients. For this reason, finding one or more
radiation oncologists interested in participation in the multidisciplinary
physician network will improve the breadth of treatment options avail-
able at a given institution. Using an immobilization device, respiratory
variation is limited during treatment. SBRT is noninvasive, delivered in the
outpatient setting, and very well tolerated. It has been studied in single
lesions up to 6 cm or up to three tumors, none greater than 3 cm. There
must be at least 700 mL of liver volume outside the treatment field. The
phase I data escalated dosage up to 16 Gy in three fractions. For the Child
A patients, there was no dose-limiting toxicity. In Child B patients, dose-
limiting toxicities were encountered and the protocol changed to a pro-
tracted 5-fraction course with the same total dose. This diminished the
toxicity to levels seen in Child A patients. The phase II data from the same
cohort of patients showed a 2-year tumor control rate of 90%. In patients
with larger tumors treated off protocol, the response rates are still excel-
lent although the long-term control rate is lower with increasing tumor
size (unpublished data).
Ablation
Ablation is a potentially curative treatment option for patients with
early-stage disease. Depending on institutional infrastructure, ablative
techniques will reside with either the interventional radiologists or the
abdominal imagers. The success of ablation is highest with lesions less
002100649.INDD 238 7/14/2014 7:19:29 PM

than 2 to 3 cm and decreases significantly in tumors larger than 3 cm.

For larger or pauci-focal tumors, ablation may be performed in combina-
tion with embolization. In solitary tumors up to 7 cm, this combination
has shown to provide 5-year survival on par with surgical resection. Both
thermal ablation (RFA, microwave, etc.) and chemical ablation (ethanol,
acetic acid) have been used to treat HCC. HCC is the ideal target for abla-
tion because it is a soft tumor surrounded by a fibrotic liver in most cases.
This is the source of the so-called “oven effect” where heat applied to the
tumor is insulated by the cirrhotic liver. The soft tumor–hard liver combi-
nation is also beneficial in chemical ablation, because the ethanol or acetic
acid can diffuse easily in the soft tumor, but is kept from escape by the cir-
rhotic liver. Which ablative technique is appropriate depends on the loca-
tion and size of the target tumor. For example, radiofrequency ablation is
susceptible to the “heat-sink” effect, in which large vessels close to tumor
can take the heat away in flowing blood and prevent complete ablation.
For a tumor close to a large vessel, microwave ablation may be a better
choice because it is not as susceptible to this effect. The long-term efficacy
for either t echnique drops substantially with increasing size and number
of lesions.
Chemotherapy
Sorafenib is FDA approved for the treatment of HCC. Since its approval,
there has been a surge in the number of HCC patients being treated with
the drug regardless of tumor stage. The use of sorafenib is based on phase II
and phase III data in patients with advanced metastatic HCC; the treatment
group showing close to a 3-month survival advantage over the nontreated
group. The objective response rate rests at around 2% with most of the
effect associated with the 35% to 71% stable disease rate noted in the phase
II and phase III trials, respectively. Over 80% of the patients in the phase
III study had been previously treated with LDTs (chemoembolization) prior
to entry. The response rate to LDTs remains above 70%, and therefore,
sorafenib must be considered in the context of all treatment options cur-
rently available. Sorafenib has been used in combination with LDTs with
reasonable toxicity profiles and slight improvement in efficacy. Dose delays
and/or reduction have been required in the vast majority of patients. Recent
phase III data investigating the benefit of sorafenib in the adjuvant setting
after embolization are less convincing. Short of a small series, sorafenib
has not been studied in the neoadjuvant setting before LDT, resection, or
transplant. Using lessons learned from other antiangiogenic compounds
used in the neoadjuvant setting, this introduces potential periprocedural
or perioperative complications that would compromise either the ability
to deliver therapy successfully or patient/graft survival. For example, with
catheter-based techniques, the arterial pruning associated with antiangio-
genic agents may impact the delivery of the small micron particle into the
tumor bed.
Cholangiocarcinoma
Cholangiocarcinoma (CCA) is commonly defined by the location of the
tumor into intrahepatic CCA, hilar CCA, or distal. The treatment of CCA has
historically hinged with surgical oncology, medical oncology, and radiation
oncology, and treatment paradigms rested mainly on whether the patient
was deemed resectable. The definition of resectable often rested primarily
with the surgeon. As nonsurgical LDTs have emerged and liver transplant
became an option for highly select patients, the importance of multidisci-
plinary input grew as did the use of neoadjuvant protocols of chemotherapy
and/or combination chemoradiation.
0002100649.INDD 239 7/14/2014 7:19:29 PM

Liver Transplant
The rationale for liver transplant for CCA centers on the relationship
between margins of resection and outcome. In hilar CCA, the intrahepatic
bile duct margin of resection is close in the majority of cases and local
recurrence most often drives outcomes. This has established medical oncol-
ogy as a more central figure in the multidisciplinary management of HCC.
With the overall response rate low, medical oncology will also be pressed
to contribute clinical trials of novel systemic agents. Most centers provid-
ing transplant as an option for these patients will have surgical oncology,
medical oncology, radiation oncology, hepatology, and organ transplanta-
tion providing some aspect of patient care. Some form of neoadjuvant treat-
ment is required prior to transplant consideration, and most centers will
follow the Mayo Clinic protocol closely, including some method to docu-
ment that lymph nodes in the porta hepatis are not involved. The natural
history of intrahepatic CCA is not well defined, and therefore, the role of
liver transplant is currently unknown.
Yttrium-90 Microspheres
Unresectable patients with intrahepatic CCA were often treated with
external beam radiation therapy and radiation-sensitizing doses of chemo-
therapy. With the familiarity of Y90 microspheres (radioembolotherapy) in
HCC, many centers have initiated protocols using combination Y90 with
gemcitabine for intrahepatic CCA with reasonable local control rates. The
theoretical advantage of this approach is improved delivery of the radioac-
tivity into the tumor bed and diminished parenchymal toxicity. There are
no established “best practices,” and the use of this approach is best served
under the direction of a clinical trial.
SECONDARY MALIGNANCIES INVOLVING THE LIVER

Many of the techniques described above have been integrated, to some
extent, into the treatment of secondary cancers such as metastatic colorec-
tal cancer. Unlike HCC, we are still trying to define where these techniques
fit into established standards of care. The following is a description of how
our multidisciplinary liver tumor group has approached their use in meta-
static colorectal cancer.
Chemotherapy
Any decision about LDT must take into account the fact that the standard of
care for metastatic colorectal cancer remains a complete course of systemic
chemotherapy. Most of the improvements in survival in this disease have been
linked to the introduction of more effective systemic agents. As patients prog-
ress from first-line treatment to second-line systemic options, the response
rate to chemotherapy drops substantially. For this reason, I feel that LDTs may
play a role either alone or in combination with second-line chemotherapy in
patients with stable disease or progression after first-line treatment.
Resection
There are ample data in large retrospective and smaller prospective series on
liver resection for metastatic colorectal cancer. Memorial Sloan Kettering
published a nomogram that will predict the 5-year survival depending on
one or more predictors. This nomogram may be useful to gauge the risk/
benefit of resection for any given patient. Survival benefit with resection is
limited to patients with liver-only disease in whom a margin-negative resec-
tion is possible and in whom all measurable disease is removed in a single
operation. Small single-institution series have suggested a role for staged
002100649.INDD 240 7/14/2014 7:19:29 PM

bilobar liver resections in select patients; however, repeat hepatectomy is

associated with higher than average liver morbidity (liver insufficiency or
failure) and diminished ability to tolerate additional systemic therapy for
extrahepatic progression postresection. With the emergence of LDTs for
secondary cancers, a clear potential advantage lies in buying potential time
before committing to resection, thereby helping select patients for surgery
who are most likely to benefit long term.
Stereotactic Body Radiation Therapy

The data using SBRT in secondary cancers, including metastatic colorectal
cancer, are quite good. The multi-institutional data in secondary cancers from
a number of primary organ sites (lung, colon, breast) showed a 93% 2-year
tumor control rate. Single-institution prospective data in a larger cohort of
patients have corroborated that result. In patients with colorectal lesions
less than 3 cm, the 2-year tumor control rate approaches 100%. The eligibility
criteria for these trials were the same as for HCC: solitary lesions up to 6 cm
or up to three lesions (none larger than 3 cm). SBRT is an excellent choice in
patients with unpredictable disease (synchronous metastatic disease) or sig-
nificant comorbid conditions making resection more risky. Again, its advan-
tage lies in buying considerable time to determine the inherent cancer biology.
Yttrium-90
Y90 is also FDA approved for metastatic colorectal cancer. The response
rates are much lower than with HCC (40% vs. 70%). Some of this decreased
response is most likely due to the tumors being less vascular. However, the
response rate of 40% is higher than the established response rate to second-
line systemic chemotherapy. With current pressure to justify cost of treat-
ment and define objective measures of benefit, clarifying the role of Yttrium
90 in metastatic colorectal cancer will become increasing more important,
more a multidisciplinary liver oncology group. For patients in whom sur-
gical resection is less likely to improve survival (multifocal and/or bilobar
disease), clinical trial data will define how Y90 therapy integrates into treat-
ment paradigms for secondary colorectal cancer.
Drug-Eluting Beads
The DEB used in metastatic colorectal cancer includes irinotecan. The elu-
tion kinetics are quite variable and, as such, the toxicity is variable as well.
Our experience with irinotecan-eluting beads has been subpar, compared
with Y90, and for this reason, Y90 remains our platform of choice.
Ablation
The results of radiofrequency ablation or microwave ablation in treating
hepatic colorectal cancer metastases are disappointing when compared
to SBRT. However, when used in combination with resection, ablation can
help eliminate all measurable disease in select patients. When used percu-
taneously under image guidance, the recurrence rate in the treated area is
moderately high (20% to 30%).
CONCLUSION
A strong, multidisciplinary panel of invested physicians from HPB surgery,
transplant surgery, hepatology, interventional radiology, radiation oncol-
ogy, medical oncology, nuclear imaging, and abdominal imaging is ideal
to develop, standardize, and continuously evaluate institutional treatment
strategies for primary and secondary liver malignancy.
0002100649.INDD 241 7/14/2014 7:19:29 PM

Suggested Readings
Andolino DL, Johnson CS, Maluccio M, et al. Stereotactic body radiotherapy for pri-
mary hepatocellular carcinoma. Int J Radiat Oncol Biol Phys 2011;81(4):447.
The Cancer of the Liver Italian Program (CLIP) Investigators. A new prognostic system
for hepatocellular carcinoma: a retrospective study of 435 patients. Hepatology
1998;28:751–755.
Cardenes HR. Price TR, Perkins SM, et al. Phase I feasibility trial of stereotactic
radiation therapy for primary hepatocellular carcinoma. Clin Transl Oncol
2010;12(3):218.
European Association for Study of Liver. EASL-EORTC clinical practice guidelines:
management of hepatocellular carcinoma. Eur J Cancer 2012;48(5):599.
Kudo M. Diagnostic imaging of hepatocellular carcinoma: recent progress. Oncology
2011;81(Suppl 1):73–85.
Llovet JM, Bru C, Bruix J. Prognosis of hepatocellular carcinoma: the BCLC staging
classification. Semin Liver Dis 1999;19(3):329–338.
Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib in advanced hepatocellular c arcinoma.
N Engl J Med 2008;359:378.
Maluccio M, Covey AM, Gandhi R, et al. Comparison of survival rates after bland arte-
rial embolization and ablation versus surgical resection for treating solitary hepa-
tocellular carcinoma up to 7 cm. J Vasc Interv Radiol 2005;16(7):955–961.
Maluccio MA, Covey AM, Schubert J, et al. Transcatheter arterial embolization with
only particles for the treatment of unresectable hepatocellular carcinoma. J Vasc
Interv Radiol 2008;19(6):862–869.
Mazzaferro V, Regalia E, Doci R, et al. Liver transplantation for the treatment of small
hepatocellular carcinomas in patients with cirrhosis. NEJM 1996;334(11):693–699.
Salem R, Lewandowski RJ, Kulik L, et al. Radioembolization results in longer time to-
progression and reduced toxicity compared with chemoembolization in patients
with hepatocellular carcinoma. Gastroenterology 2011;140(2):497.
Shiina S, Tateishi R, Arano T, et al. Radiofrequency ablation for hepatocellular
carcinoma: 10-year outcome and prognostic factors. Am J Gastroenterol 2012;
107(4):569–577.
Yoshitaka T, Kazuhiro N. Nonalcoholic steatohepatitis-associated hepatocellular
carcinoma: Our case series and literature review. World J Gastroenterol 2010;
16(12):1436.
002100649.INDD 242 7/14/2014 7:19:29 PM

20 Laparoscopic Liver Surgery
Juan M. Sarmiento and Edward Lin
INTRODUCTION
Liver resection is the final frontier yet to be conquered by laparoscopy. In
spite of the fact that it has been 20 years since the description of the first
laparoscopic liver resection (nonanatomic wedge resection), there is no
universal application of the laparoscopic technique. In fact, laparoscopic
major hepatectomy is still done sparsely around the world and in the
United States. There are some explanations for the lack of enthusiasm for
adopting this technique. First, some of the maneuvers currently utilized in
open surgery are not available laparoscopically (mobilization, palpation,
compression, etc.) with the consequent fear of torrential or uncontrollable
bleeding. Second is the continuous fear of air (gas) embolism with hemo-
dynamic disturbances, including cardiac collapse. Third, tumors located
near the major vessels or biliary structures. Finally a natural learning curve
that is less tolerant to surgical mishaps due to the potentially catastrophic
consequences of hemorrhage. Fear of oncologic in adequacy of the laparo-
scopic technique is less of an issue nowadays after the publication of the
COST trial for laparoscopic colectomy.
Up until 2009, there have been about 3,000 reported cases of lapa-
roscopic liver operations, most of them minor resections (<3 segments).
Still, the morbidity (10%) and mortality (<1%) rates of such series remain
low. These numbers compare favorably with the open approach and call
for a renewed enthusiasm on the application of laparoscopic techniques
to liver surgery. Even though the use of this relatively new technique is
somewhat in its infancy, and therefore a scientifically proven evalua-
tion is in order, the hope for a randomized trial is quickly fading. Those
who think laparoscopy offers real value consider it unethical to submit
patients to a more aggressive form of therapy, and those who are not
laparoscopic experts do not have the skills to mount a comparable series
to their open technique experience. Therefore, the knowledge we can
acquire in this situation is driven from retrospective studies and from
relatively small series coming from major centers with experienced lapa-
roscopic surgeons.
Another limitation to the use of laparoscopic hepatectomy has to do
with surgeon training. A classically trained hepatobiliary surgeon ( from
any track) is focused on open resections with exposure to operations done
through large incisions. On the other hand, the laparoscopic expert is sel-
dom trained in liver resections, as the focus of their technical expertise is
hollow viscus, bariatric surgery, and small solid organs such as the adre-
nal glands and spleen. The point of confluence remains to be defined as
the practice setting is also completely different and the referral of patients
limited. The best route to overcome such limitation is the fusion of both
disciplines and the intimate collaboration by those surgeons coming from
their respective tracks.
243
0002086386.INDD 243 7/14/2014 7:41:04 PM

By 2007, the majority of the experience came from small case series of
mostly minor liver resections. Koffron in 2007 presented the largest series
then from a single institution of 300 cases, with a very good number of
major hepatectomies. This series is probably responsible for changing the
landscape on the feasibility and application of this technique.
The need to standardize this technique came to realization in
2008 after the international conference on laparoscopic liver surgery in
Louisville, KY. This conference brought together expert hepatobiliary sur-
geons, both with and without experience in laparoscopic liver resection,
and produced a document with the following recommendations:
■■ The group cautioned on the expansion of the application of the lapa-
roscopic technique for benign lesions that were not offered resection
otherwise.
■■ There was a concern for patient safety due to a rapid and indiscriminate
growth of programs offering this technique without adequate training of
surgeons and paramedical personnel and minimal standards of quality.
■■ There was clarity in defining laparoscopic surgery (“pure laparoscopy”)
and hand-assisted ( for obvious reasons) and hybrid technique (laparos-
copy for mobilization and dissection but transection performed via an
open approach).
■■ There was a definition for the ideal candidate for laparoscopic hepatec-
tomy: patients with solitary small (<5 cm) lesions, located peripherally
(segments II through VI). Also, it was stated that left lateral sectionec-
tomy should be done laparoscopically as the standard of practice.
■■ Need for expeditious recognition of conversion to an open approach due
to safety or lack of progress.
■■ Need for a cooperative registry for all laparoscopic liver resections to
monitor outcomes, especially in high-risk resections (major hepatec-
tomy, living donors). A randomized trial was probably impractical, and
this registry would serve as an alternate source of information and
follow-up.
■■ Definition that resection, either open or laparoscopic, is the gold stan-
dard for the treatment of colorectal metastases. Caution was given to the
need of keeping comparable rates of negative-margin resections and the
avoidance of missing occult lesions.
■■ Definition of how laparoscopic resection is still a valuable tool in the
treatment of hepatocellular carcinoma (HCC), and no specific distinc-
tion was made between ablation and transplantation.
One of the realities in dealing with laparoscopic liver resections
is the great variance of operations fitting the definition. If we somewhat
simplify the term, and call it major resection, we can describe at least five
operations that fit this term: right lobectomy, extended right lobectomy, left
lobectomy, extended left lobectomy, and central hepatectomy. If we add the
caudate lobe, technically a minor liver resection but operatively demand-
ing, and also add combined liver and biliary resection, there is a whole
spectrum of “major” resections to be mastered laparoscopically. Other sub-
specialized areas like Whipple procedures or colectomies really encompass
two to three operations that are more reproducible and amenable to stan-
dardization. Like any other operation, the surgeons’ comfort grows with
experience in the procedure; laparoscopic wedge resections are tackled at
first, then progressing to the laparoscopic left lateral sectionectomy, and
then to subsegmentectomy/segmentectomy II to VI. With the progressive
experience, the confidence of the surgeon and dexterity will increase, and
at some point, the surgeon will be ready to take on laparoscopic major
hepatectomy, resembling the learning process of the open liver resection.
002086386.INDD 244 7/14/2014 7:41:04 PM

Chapter 20 / Laparoscopic Liver Surgery 245
The hand-assisted technique could even bridge that transition as it applies

principles of both the open and the laparoscopic approach. The hand is par-
ticularly useful for the mobilization of the liver, the palpation of structures,
the compression of the liver parenchyma for hemostatic purposes, as well
as the rapid and secure control of bleeding from major vascular structures.
In addition, the hand port is used for specimen extraction, an incision that
is unavoidable for major liver resection. Along the same lines, rapid conver-
sion to an open approach early in the experience and implementation of the
hybrid technique are also valid strategies to reach the degree of confidence
necessary to achieve competence in laparoscopic liver resections.
As Cherqui points out, laparoscopic liver resection should be based upon
the foundation of open liver surgery, being that laparoscopy is most difficult
of the two major skill sets necessary for this technique. The surgeon collabo-
ration is mutually enriching and, based on the maturity of the surgeons, will
serve as the basis for independent criticism of their own technique and the
development of proficiency and standardization of these cases. Doubling the
experience to two surgeons with a single case is another perk in a field with
limited number of cases. Finally, the application of production quality mea-
surement tools adopted from the manufacturing industry has helped with the
refinement of the final product, in this case laparoscopic hepatectomy.
LAPAROSCOPIC MAJOR HEPATECTOMY

Dagher et al. published an international registry study collecting informa-
tion on 210 patients undergoing laparoscopic major hepatectomy, 57%
being hand-assisted approaches. They reported a mortality of 1% and a
morbidity of 8% with a conversion rate of 12%. Experience over time showed
that operative time, blood loss, portal triad clamping time, conversion
rate, and length of stay were improved. Martin et al. from the University of
Louisville more recently showed in a matched study of laparoscopic major
liver resection a mortality of 1% with a morbidity of 23% and a conversion
rate of 4%. Blood loss, use of the Pringle maneuver, length of stay, compli-
cation rate, operative time, and blood transfusion were significantly better
than the matched open group. These two studies show that laparoscopic
major liver resection can be achieved with acceptable results in terms of
complications and, that with enough experience, all the feared events (i.e.,
hemorrhage, increased operative time, need for transfusion) not only are
manageable but result in better outcomes.
Two recent meta-analyses (including both minor and major laparo-
scopic liver resections) have shown similar results to the two studies men-
tioned above, with the exception that operative time was longer with the
laparoscopic method. However, blood loss, morbidity rate, and length of
stay were significantly better for the laparoscopic group with similar mar-
gin-negative rates. Taking into account the inherent bias in the selection of
patients submitted to laparoscopic resection, there is at least a comparable
outcome profile for the laparoscopic technique, with a trend toward lower
blood loss and decreased hospital stay.
ONCOLOGIC APPLICATIONS
Colorectal Metastases
The fact that no randomized trial has been done for liver resection as the
treatment of colorectal metastases basically defines the impossibility of
doing one for laparoscopic versus open resection. After the COST trial for
surgical resection of colorectal cancer, where no oncologic differences were
found between the two approaches (laparoscopic vs. open), there is suffi-
cient approval of laparoscopy as a technique for resection, if feasible, for
0002086386.INDD 245 7/14/2014 7:41:05 PM

colorectal cancer. The other issue is that most of these patients have prior
abdominal operations (colectomy) and adhesions could be difficult to over-
come for a definitive liver resection. An international multicenter study of
109 patients was reported by Nguyen et al. in 2009. Almost 40% of the resec-
tions were a full lobectomy; 95% of patients had negative margins with a
5-year disease-free survival of 43% and an overall survival at 5-year of 50%.
Another study from Europe of about 150 patients showed a 93% margin-
negative rate, though major resection accounted for less than 20% of the
patients in that series. Although the follow-up was short, the disease-free
survival was similar to that seen in open resection. The above-cited studies
were descriptive in nature and basically, they both justify per se the lapa-
roscopic approach, when feasible, for resection of colorectal metastases to
the liver. Another important article from France was done with case–con-
trol methodology, including 60 patients in each group and having compara-
ble incidence of major hepatectomy (about 40% in each group). Important
differences shown in this paper include a higher R0 resection rate (87% vs.
72%) in the laparoscopic group. In terms of survival, the 5-year recurrence
free rate was similar (35% laparoscopic vs. 27% open) as well as the over-
all survival (64% laparoscopic vs. 56% open). It is perfectly valid to assume
after studying all these series that the oncologic safety of the laparoscopic
procedure is acceptable and that the administration of pneumoperitoneum
does not increase the recurrence of the tumor in these patients.
Hepatocellular Carcinoma
The higher number of studies in laparoscopic resection for HCC prob-
ably speaks more of the higher incidence of this condition than the com-
fort of the surgeons with this technique. Ninety percent of HCC patients
have accompanying cirrhosis, and the implications of this association are
diverse. First, the manipulation of the liver with cirrhosis is more complex,
with a higher tendency to tear and bleed, and also the hard consistency of
the fibrotic liver complicates parenchymal transaction. Second, pneumo-
peritoneum increases the abdominal pressure creating ischemia to the liver
and the kidneys, with a further decrease of glomerular filtration (a common
occurrence in cirrhosis) and possible renal decompensation. Third, portal
hypertension and coagulopathy, again a common occurrence in cirrhotic
patients, increase intraoperative bleeding and obscure the limited laparo-
scopic field; this coupled with the fact that manual compression/manip-
ulation is not readily available theoretically could alert the surgeon to be
more cautious when submitting these patients to laparoscopy. In spite of
the above considerations, there are more data on HCC than on colorectal
metastases, although for obvious reasons, the rate of major hepatectomy
is consistently low for the HCC series. Two matched studies (Belli et al.
and Sarpel et al.) have shown comparable results between laparoscopic
and open techniques for resection of HCC. The first one shows a similar
mortality rate but a lower morbidity rate for the laparoscopic approach.
Interestingly, overall survival at 3 years was similar (67% lap, 62% open) as
well as the disease-free survival (52% lap, 59% open). Furthermore, the sur-
vival statistics are in accordance with traditional series on survival for the
open approach. The second matched study confirmed the similar rates of
overall survival and disease-free survival as the first study, without differ-
ence in perioperative results but a lower hospital stay favoring the laparo-
scopic group. Hence, matched studies show again the oncologic safety of
the laparoscopic approach.
A meta-analysis of nine studies performed by Fancellu in 2011 com-
piled 600 patients, 40% of them performed laparoscopically. Among the
differences found, the laparoscopic group had a higher R0 resection rate,
002086386.INDD 246 7/14/2014 7:41:05 PM

shorter hospital stay, lower operative blood loss and need for transfusions,
lower rates of liver failure and ascites, and a trend for lower perioperative
mortality without increase in operative time.
A last point is worth mentioning: When the liver resection is used as a
bridge to transplantation or the patient needs a rescue liver transplant, the
laparoscopic approach has proved to create fewer adhesions and decreases
the complexity of the procedure as demonstrated by lower operative time,
blood loss, and need for transfusion during the hepatectomy phase of the
liver transplant.
Overall, all the results of the laparoscopic technique are at least
comparable to the open approach in managing HCC and do not alter the
recurrence and survival rates compared to open resection. Obviously, these
operations are more challenging due to the reasons explained above but
experienced laparoscopic HPB surgeons can approach these patients with
confidence in both perioperative and long-term results.
ANATOMIC RESECTIONS
In an effort to make comparisons more realistic, as all the series mentioned
above contain a great range of the type of liver resections, it is important
to include in this analysis a direct comparison of patients undergoing both
techniques for the same type of hepatectomy.
Left Lateral Sectionectomy

In an analysis by Cherqui’s group in 2003, which includes patients with any
type of disease, operative blood loss and hospital stay were lower in the lapa-
roscopic group, with no differences in morbidity but an increase in operative
time. This study was presented early in the development of the laparoscopic
technique. Another study performed at Pittsburgh compared 29 laparo-
scopic cases with 40 open; again, the length of stay was lower in the lapa-
roscopic group, and the morbidity rate was also lower. Interestingly, adding
these two factors to an economic model, they determined that laparoscopy
saved between 1,500 and 3,000 US dollars per patient. Overall, as determined
by the international conference in 2008, this specific type of liver resection
should now be approached laparoscopically as the primary option.
Right Hepatectomy
This is the most common major hepatectomy done, and it is probably the
best model to establish the standard results and also to compare with the
open approach. Right hepatectomy is very well defined traditionally, is done
routinely in major centers, and is the most common major resection of the
liver. On the other hand, as surgeons become facile with the laparoscopic
approach while other experienced members of the team continue perform-
ing the traditional open approach, this creates a setting for comparative
studies over the same time period. The two largest published studies rep-
resent that scenario. Even in that setting, laparoscopic right hepatectomy
accounted for 22 patients in a French study and 36 in another. In the first
study by Dagher et al., patients were matched by demographics and risk
factors (comorbidities). The results showed similarity in operative times
(an improvement, as traditionally laparoscopic cases take longer) and liver-
specific complications, but lower blood loss, shorter hospital stay, and lower
overall morbidity rate for laparoscopy. In the second paper by Hilal et al.,
also case-matched, operative times were longer in the laparoscopic group,
with similar morbidity rates and blood loss, but with a significantly shorter
hospital stay. The authors of this review also have shown with a comparative
study of about 50 laparoscopic right hepatectomies that all the perioperative
0002086386.INDD 247 7/14/2014 7:41:05 PM

parameters are improved including operative time (unpublished data). Also,

we found a significantly lower morbidity rate and a significantly shorter
hospital stay for the laparoscopic cases. Even though operative costs were
increased, they were compensated by early dismissal and lower complica-
tions resulting in a similar overall cost for both procedures.
Laparoscopic right hepatectomy should also serve as a tool to estab-
lish standards of quality for major HPB centers. This operation is easily
deconstructed in steps, amenable to measurement (each one of them), and
is done frequently enough to produce data from each center. We have done
this structured comparison and have found consistency for each of the
steps (i.e., low standard deviation) except for the parenchymal transaction.
We have observed that liver-specific features such as size and presence of
steatosis or cirrhosis (all of them beyond the surgeon’s control) account for
some of these variances. By defining these steps and identifying the vari-
ance, we can measure costs associated with these differences and predict
complication rates and associated costs.
Overall, laparoscopic right hepatectomy and, by extension, major
hepatectomy in general are associated with shorter hospital stay and com-
parable blood loss/transfusion rates. The morbidity is probably lower as
well, and operative times, which tend to improve with experience and stan-
dardized protocols, will be amenable to improvement with this technique.
We predict that laparoscopic major hepatectomy will become as routine
as left lateral sectionectomy is today and certainly could potentially be the
main indication for laparoscopic liver resection.
Other Major Resections

There is interest for other types of anatomic resections but with sparse data.
Living donor left lateral sectionectomy was reported in the last decade for
17 cases with overall good results. Some data have been included also on
donor right hepatectomy (this one mostly with the hybrid approach), again
with comparable results to the open technique. The authors of this chap-
ter have presented a small series of caudate resection and combined major
hepatectomy with bile duct resection/reconstruction, which are techni-
cally demanding but are almost unexplored in the field of laparoscopic
resections. These patients so far have had good results and show promise as
this technique is further developed.
ROBOTIC LIVER SURGERY

Robotic surgery is an extension of minimally invasive liver resection.
Through the experience in prostatectomy and other abdominal opera-
tions, there has been a proven benefit with emphasis on shorter hospital
stay, better cosmetic results, and lower postoperative pain. Advantages of
tridimensional view (resembling the human eye), greater articulation, and
overall resemblance of the surgeon’s physical approach to these opera-
tions engender enthusiasm toward applying this technique to liver surgery.
However, the learning curve, the high level of dexterity needed, and the
overall enhanced technical skills make the robotic technique less available
for widespread use. With better devices to control bleeding and approach
parenchymal transactions, some centers are starting to reexplore the
robot’s use in liver surgery. The robotic technique is so young that it was not
discussed at the international conference in 2008.
It is important to point out certain potential benefits of robotic liver
surgery. Biliary reconstructions are theoretically ideal for the dexterity and
“wrist” articulation of the robot. There are case reports of robotic Kasai
procedures, choledochal cyst resection, and biliary reconstruction. It offers
002086386.INDD 248 7/14/2014 7:41:05 PM

advantages in this setting over laparoscopy in terms of better ergonomics,

increased dexterity, and higher degree of needle rotation. This improve-
ment for biliary anastomosis is also applied to combined liver/biliary resec-
tions, even if the liver resection only uses laparoscopy and not the robot.
The addition of the third robot arm is critical here to set-up the anastomo-
sis, to keep gentle retraction (i.e., the base of segment IV), and to facilitate
the formation of the anastomosis or even facilitate hemostasis.
There is also interest in combined robotic abdominal resections. Since
rectal surgery is really going in the direction of robotics, and about 20% of
the patients present with simultaneous liver metastases, there is a renewed
interest in combining these two resections robotically. Like in any other
cancer operation, the hope with this technique is to improve the recovery
of the patient and make the patient fit for timely start of chemotherapy.
There is no question that robotic liver surgery is in its infancy. Most
series are very low in number and mostly done by single surgeons who
developed focused interest in this practice. The difficult learning curve and
the challenge of new instruments will select the surgeons who want to take
advantage of the robot platform. Even after conquering the laparoscopic
phase of liver surgery, few surgeons will be prepared to advance to the robotic
phase. We must remember that there was also fear that came at the beginning
of the laparoscopic era, but laparoscopy is the standard of practice for many
abdominal procedures nowadays. With the improved ergonomics of the
robot and the incorporation of haemostatic and transecting devices to the
robot platform, the challenges of hepatic surgery are ready to be conquered.
TECHNIQUE OF LAPAROSCOPIC LIVER RESECTION

This technique should follow both the indications and the initial steps of
the open procedure. The surgeon should master laparoscopic intraopera-
tive ultrasound and should follow the initial inspection of the abdominal
cavity as the first stages of the operation. For wedge resections, the surgeon
should triangulate as three trocars usually suffice: the camera being placed
in direct (“perpendicular”) view of the segment to resect and the other two
ports at a variable distance that allows for freedom of movement with com-
fortable distance to the target area. Most wedges can be done with regular
hemostatic devices like bipolar energy, ultrasonic scalpel, and radiofre-
quency ablation. Use of staplers is not necessary for the most part.
Laparoscopic left lateral sectionectomy is very straightforward. The
camera is placed in the umbilicus or slightly higher. Then the other two
ports are placed above the level of the umbilicus, close to the midclavicular
line (Fig. 20.1). We use 10- to 12-mm trocars in all the ports to exchange sta-
pling devices and large graspers on both sides of the midline. The transec-
tion is started to the left of the umbilical fissure with any hemostatic device
after “hanging” the liver from the round ligament (approached from the
right trocar). After thinning out the liver parenchyma and identifying the
portal pedicles, an endovascular stapler is fired from the left trocar; usually
two loads are necessary to take both the pedicle for segment 3 and the one
for segment 2. The rest of the transection is done with the energy device,
and care must be taken in the posterior area to avoid an injury to the left
hepatic vein. Upon identification of the left hepatic vein, dissection is done
around it, and then, the stapler is fired this time from the right-sided trocar.
Then, parenchymal transaction is completed, and the triangular ligament is
transected as the last step of the resection. The specimen is removed via an
extended incision over the umbilical trocar.
Laparoscopic Right Hepatectomy. We follow the same technique as the
open approach. The advantages of laparoscopy are better visualization,
0002086386.INDD 249 7/14/2014 7:41:05 PM

FIGURE 20.1 Trocar placement for left lateral sectionectomy.
especially of the areas behind the liver, over the inferior vena cava (IVC),
over the right hepatic vein, etc. Also, the increased pressure of the pneu-
moperitoneum keeps the operative field relatively bloodless as small
venules on the transection line are collapsed. Since the patient will need
an incision for specimen extraction, a hand-assisted technique has been
developed with an additional advantage for liver mobilization, compres-
sion, and urgent management of major vascular bleeding. We place the
trocars as shown in Figure 20.2. No mobilization of the liver is done after
the ultrasound, and the gallbladder is kept in situ for traction. The right
hepatic artery is visualized after taking the cystic artery and is ligated and
transected. The right portal vein is transected with endovascular staplers.
This is important as the space to the area of the right portal pedicle “opens
up” with traction and moves away allowing for a more lateral transection,
thus avoiding injury to the contralateral bile duct (or the common hepatic
duct). The right bile duct is transected within the liver. Mobilization is
done all the way up to the right hepatic vein, and then we use energy
devices to take all the small vessels of the anterior surface of the IVC until
identification of the IVC ligament, which is taken with staplers. After this,
the right hepatic vein is dissected and stapled. With the line of demarca-
tion now present, we score the capsule of the liver and then start tran-
secting with energy devices, which is done blindly for the first 2 to 3 cm
in depth as the vessels lying in that area are within the hemostatic capa-
bilities of these devices. With further dissection, we take smaller bites on
002086386.INDD 250 7/14/2014 7:41:05 PM

FIGURE 20.2 Trocar placement for laparoscopic right hepatectomy.
the transection area, making sure not to injure major tributaries of the
hepatic veins. These will be taken with staplers. When the transection is
finished, the specimen is removed through the hand port and pneumo-
peritoneum is reinstated to secure hemostasis and repair any bile leaks.
No drains are placed.
We have found that the Pringle maneuver is not necessary for any of
these major hepatectomies and that the blood loss is minimal. In fact, our
blood transfusion rate is less than 5%.
Laparoscopic liver resection is a technique that will be used more fre-
quently in the future. Established advantages like decreased blood loss and
shorter hospital stay will make it even more attractive in the era of cost con-
tainment and bundled payments. Other possible advantages like decreased
complication rates with similar operative times will put this technique even
more in the forefront of the liver surgery armamentarium. Robotic liver sur-
gery is still very undeveloped, but it has great promise not only to achieve
similar results to laparoscopy but also to overcome the limitations of the
latter technique.
Suggested Readings
Buell JF, Cherqui D, Geller DA, et al. The International position on laparoscopic liver
surgery. The Louisville statement, 2008. Ann Surg 2009;250:825–830.
Dagher I, Belli G, Fantini C, et al. Laparoscopic hepatectomy for hepatocellular carci-
noma: a European experience. J Am Coll Surg 2010;211:16–23.
0002086386.INDD 251 7/14/2014 7:41:06 PM

Kluger MD, Vigano L, Barroso R, et al. The learning curve in laparoscopic major liver
resection. J Hepatobiliary Pancreat Sci 2013;20:131–136.
Koffron AJ, Auffenberg G, Kung R, et al. Evaluation of 300 minimally invasive liver
resections at a single institution. Less is more. Ann Surg 2007;246:385–394.
Nguyen KT, Gamblin TC, Geller DA. World review of laparoscopic liver resection.
Ann Surg 2009;250:831–841.
Nguyen KT, Laurent A, Dagher I, et al. Minimally invasive liver resection for metastatic
colorectal cancer. Ann Surg 2009;250:842–848.
002086386.INDD 252 7/14/2014 7:41:06 PM

SECTION III
Biliary
0002086387.INDD 253 7/14/2014 7:43:57 PM

21 Liver and Biliary Anatomy
Steven M. Strasberg
This chapter is an overview of hepatobiliary anatomy. It covers the prevailing,

that is, commonest anatomic patterns and common anomalies, with empha-
sis on surgically important anomalies.
ANATOMY AND NOMENCLATURE OF THE LIVER

The Brisbane 2000 Terminology of Hepatic Anatomy and Resections used
throughout this chapter is the official nomenclature adopted by the
International Hepato-Pancreato-Biliary Association.
Ramification of the Hepatic Artery

The primary ( first-order) division of the proper hepatic artery is into the
right and left hepatic arteries (Fig. 21.1). These branches supply arterial
inflow to the right and left hemilivers or livers (Fig. 21.2). The border or water-
shed of the first-order division is the midplane of the liver. It intersects the
gallbladder fossa and groove for the inferior vena cava (IVC) (Fig. 21.2). The
right liver usually has a larger volume than the left liver (60:40), although
this is variable.
The second-order branches (Figs. 21.1 and 21.2) of the hepatic artery
supply the four hepatic sections. The right liver has two sections, the right
anterior section and the right posterior section, supplied by the right anterior
sectional hepatic artery and the right posterior sectional hepatic artery,
respectively (Fig. 21.1). The plane between these sections is the right inter-
sectional plane, which does not have surface markings to indicate its posi-
tion (Fig. 21.2). The left liver also has two sections, the left medial section
and the left lateral section, which are supplied by the left medial sectional
hepatic artery and the left lateral sectional hepatic artery (Fig. 21.1). The
plane between these sections is referred to as the left intersectional plane.
It does have surface markings indicating its position—the umbilical fissure
and the line of attachment of the falciform ligament to the anterior surface
of the liver.
The third-order branches of the hepatic artery divide the right
and left hemilivers into segments 2 to 8 (Figs. 21.1 and 21.2). Each of
the segments has its own feeding segmental artery. The left lateral sec-
tion is divided into Sg2 and Sg3. The pattern or ramification of vessels
within the left medial section does not permit subdivision of this sec-
tion into segments, each with its own arterial blood supply. Therefore,
the left medial section and Sg4 are synonymous. However, segment 4 is
arbitrarily divided into superior (4a) and inferior (4b) parts. The right
anterior section is divided into two segments, Sg5 and Sg8. The right
posterior section is divided into Sg6 and Sg7. The planes between seg-
ments are referred to as intersegmental planes. The ramifications of the
bile ducts are identical to that described for the arteries as are the zones of
the liver drained by their respective ducts.
254
002086387.INDD 254 7/14/2014 7:43:57 PM

Chapter 21 / Liver and Biliary Anatomy 255
FIGURE 21.1 Three orders of ramification of the proper hepatic artery are shown color
coded. First order (red) is division into right (A) and left (B) hepatic arteries. The second
order is division into sectional arteries (green), including the right anterior (c), the right
posterior (d), the left medial (e), and the left lateral (f) sectional arteries. The third-order
division (blue) is into segmental arteries that are numbered and correspond to the Couinaud
segments. The three orders supply the hemilivers or livers, sections, and segments. Note
that the second order and third order for the left medial sectional artery and artery to seg-
ment 4 are identical (banded green and blue). Segment 1, which is separate from the two
hemilivers, is supplied by the arteries that arise from the right and left hepatic arteries (not
shown). Ramification of the bile ducts is identical to that of the arteries. (From Strasberg
SM, Philips C. Use and dissemination of the Brisbane 2000 nomenclature of liver anatomy
and resections. Ann Surg 2013;257(3):377–382.)
Segment 1 (caudate lobe) is a distinct portion of the liver, separate

from the right and left hemilivers. It has three parts: the bulbous left part
(spigelian lobe); the paracaval portion, which lies anterior to the vena cava;
and the caudate process (or right part), on the right (Fig. 21.3). The cau-
date lobe is situated posterior to the hilum and the portal veins. The cau-
date receives vascular supply from both right and left hepatic arteries (and
portal veins). Caudate bile ducts drain into both the right and left hepatic
ducts. The caudate lobe is drained by several short caudate veins that enter
the IVC directly. These veins enter the IVC on either side of the midplane
of the vessel, an anatomical feature that normally allows the creation of a
tunnel behind the liver on the surface of the IVC without encountering the
caudate veins.
Resectional Terminology
Anatomic liver resections are hemihepatectomies (or hepatectomies), sectio-
nectomies, or segmentectomies. The nomenclature for specific resections is
described in Figures 21.4 to 21.8.
Resection of segment 1 is usually called a caudate lobectomy.
0002086387.INDD 255 7/14/2014 7:43:58 PM

256 Section III / Biliary
Midplane of Liver
Left Hemiliver/Liver
r
ive
/L
er
iliv
em
H
ht
ig
R
Line of insection of vessels

and bile ducts
Inferior vena cava
A First Order: The Hemilivers or Livers
Right Left
Intersectional Intersectional
Plane Plane
LMS LLS
RAS
RPS
B Second Order: The “Sections”

FIGURE 21.2 A. First-order division of the liver into the hemilivers or livers. Midplane of
the liver shown in red. B. Second-order division of the liver into sections. (RPS, right poste-
rior section; RAS, right anterior section; LMS, left medial section; LLL, left lateral section.)
The intersectional planes are shown in green.
(Continued )
002086387.INDD 256 7/14/2014 7:44:01 PM

2
8 3
7 4
6
5
1
Third Order: The “Segments”
C
FIGURE 21.2 (Continued) C. Third-order division into numbered segments. (From
Strasberg SM, Philips C. Use and dissemination of the Brisbane 2000 nomenclature of liver
anatomy and resections. Ann Surg 2013;257(3):377–382.)
SURGICAL ANATOMY FOR LIVER RESECTIONS

Hepatic Arteries and Liver Resections
In the prevailing anatomic pattern, the celiac artery terminates by divid-
ing into splenic and common hepatic arteries. The common hepatic artery
runs anteriorly and to the right to ramify into gastroduodenal and proper
hepatic arteries. The proper hepatic artery enters the hepatoduodenal liga-
ment and normally runs for 2 to 3 cm along the left side of the common
bile duct (CBD) and terminates by dividing into the right and left hepatic
arteries, the right immediately passing behind the common hepatic duct
(CHD). The four sectional arteries arise from the right and left arteries
1 to 2 cm from the liver. Variations from this pattern, such as replaced (aber-
rant) arteries, are common. “Replaced” means that the artery supplying a
particular volume of the liver is in an unusual location and also that it is the
FIGURE 21.3 Anatomy of caudate lobe. Segment 1-1,

caudate lobe proper; Segment 1-r, caudate process; MHV,
middle hepatic vein; PV, portal vein; IVC, inferior vena cava
0002086387.INDD 257 7/14/2014 7:44:04 PM

FIGURE 21.4 Resectional terminology for excision of a hemiliver or liver. (From Strasberg SM,
Philips C. Use and dissemination of the Brisbane 2000 nomenclature of liver anatomy and
resections. Ann Surg 2013;257(3):377–382.)
sole supply to that volume of the liver. This should not be confused with the
term “accessory” artery. Accessory in this sense indicates that the artery is
not the sole blood supply to an area (Fig. 21.8).
Part or all of the liver is supplied by a replaced artery in 25% of patients.
The replaced right hepatic artery arises from the superior mesenteric artery
FIGURE 21.5 Resectional terminology for excision of a section. (From Strasberg SM,
002086387.INDD 258 7/14/2014 7:44:06 PM

FIGURE 21.6 Resectional terminology for excision of a segment. (From Strasberg SM,
(SMA) and usually runs behind and then along the right posterior border of
the CBD (Fig. 21.8), where it may often be palpated. It may supply a segment
of, a section of, or the entire right hemiliver. The replaced left hepatic artery
arises from the left gastric artery and courses in the lesser omentum to the
liver. Rarely, a replaced artery supplies the entire liver, and then it is called a
replaced common hepatic artery.
Bile Ducts and Liver Resections

The prevailing pattern of bile duct drainage from the right liver is shown in
(Fig. 21.9A). There are important biliary anomalies on the right side of the
liver. The right posterior sectional duct inserts into the left hepatic duct in
FIGURE 21.7 Terminology for extended resections (three sections). (From Strasberg SM,
0002086387.INDD 259 7/14/2014 7:44:07 PM

A Normal anatomy B Replaced left hepatic artery
Gallbladder Cystic artery

Replaced left
hepatic artery
Cystic artery Hepatic duct
Portal vein
Proper Proper
hepatic artery Aorta hepatic artery
Celiac trunk Splenic artery
Gastroduodenal artery Gastroduodenal
artery
Right gastric artery Stomach Portal vein
Pylorus
Duodenum Pancreas
C Accessory left hepatic artery D Replaced right hepatic artery
Accessory
left hepatic
Proper artery Portal vein
hepatic artery Left gastric Aorta
artery Gastroduodenal
Gastroduodenal artery Splenic artery artery
Splenic artery
Replaced right
hepatic artery
Superior
mesenteric
artery
E Replaced common hepatic artery off SMA F Anterior location of right hepatic artery
Right hepatic
artery
Proper Portal vein
hepatic artery Lienogastric Aorta
trunk Celiac trunk
Gastroduodenal artery Splenic artery
Replaced Hepatomesenteric
hepatic artery trunk Stomach
Pancreas
FIGURE 21.8 Prevailing pattern (“normal anatomy”) and some common variations of the
hepatic artery. (From Mulholland MW, Lillemoe KD, Doherty GM, et al. Greenfield’s surgery:
scientific principles & practice, 5th ed. Philadelphia, PA: Lippincott Williams & Wilkins, 2010.)
20% of persons (Fig. 21.9B), and the right anterior bile duct does so in 6%
(Fig. 21.9C). A right sectional bile duct inserting into the left hepatic duct
is in danger of injury during left hepatectomy. Another important anomaly
is insertion of a right bile duct into the biliary tree at a lower level than
the prevailing site of confluence (Fig. 21.9D). The latter anomaly places the
aberrant duct at great risk for injury during laparoscopic cholecystectomy.
The prevailing pattern of bile duct drainage from the left liver and
common anomalies is shown in Figure 21.10.
Portal Veins and Liver Resections
On the right side of the liver, the portal vein divisions correspond to those of
the hepatic artery and bile duct, and they supply the same hepatic volumes
(Fig. 21.11). It divides into two sectional and four segmental veins as do the
arteries and bile ducts. The left portal vein consists of a horizontal or trans-
verse portion, which is located under Sg4, and a vertical part or umbilical por-
tion, which is situated in the umbilical fissure (Fig. 21.11). Unlike the right
portal vein, neither portion of the left portal vein actually enters the liver,
but rather, they lie directly on its surface. Often, the umbilical portion is hid-
den by a bridge of tissue passing between left medial and lateral sections.
The junction of the transverse and umbilical portions of the left portal vein
is marked by the attachment of a stout cord—the ligamentum venosum.
This structure, the remnant of the fetal ductus venosus, runs in the groove
between the left lateral section and the caudate lobe and attaches to the left
hepatic vein/IVC junction.
002086387.INDD 260 7/14/2014 7:44:12 PM

FIGURE 21.9 Prevailing pattern and important variations of bile ducts draining the right
hemiliver. (From Fischer JE, Jones DB, Pomposelli FB, et al. Fischer’s mastery of surgery, 6th ed.
FIGURE 21.10 Prevailing pattern and important variations of bile ducts draining the left
hemiliver. (From Fischer JE, Jones DB, Pomposelli FB, et al. Fischer’s mastery of surgery,
0002086387.INDD 261 7/14/2014 7:44:14 PM

FIGURE 21.11 Ramification of the portal vein in the liver. The portal vein divides into right
(A) and left (T ) branches. The branches in the right liver (the sectional branches [c] and [d ],
and the numbered segmental branches) correspond to those of the hepatic artery and bile
duct. The branching pattern on the left is unique. The left portal vein has transverse (T ) and
umbilical portions (U ). The transition point between the two parts is marked by the attach-
ment of the ligamentum venosum (L.V.). All major branches come off the umbilical portion.
The vein ends blindly in the ligamentum teres (L.T.). (Copyright Washington University in
St. Louis.)
The transverse portion of the left portal vein sends only a few small
branches to Sg4. Large branches from the portal vein to the left liver arise
exclusively beyond the attachment of the ligamentum venosum, that is,
from the umbilical part of the vein (Fig. 21.11). Branches come off both sides
of the vein—those arising from the right side pass into Sg4 and those from
the left supply Sg2 and Sg3. The left portal vein terminates in the ligamen-
tum teres at the free edge of the left liver.
The most common variation is absence of the right portal vein. In
these cases, the right posterior and right anterior sectional portal veins
originate independently from the main portal vein. A rare but poten-
tially devastating anomaly is the absent extrahepatic left portal vein, an
anomaly in which the main portal vein enters the right side of the liver
and courses intrahepatically to supply the left liver. Division of this vein
thinking that it is the right portal vein will lead to portal devascularization
of the entire liver.
Hepatic Veins and Liver Resection
Three large hepatic veins run in the midplane of the liver (middle hepatic
vein), the right intersectional plane (right hepatic vein), and the left
intersectional plane (left hepatic vein). The left hepatic vein begins in the
plane between Sg2 and Sg3 and travels in that plane for most of its length.
002086387.INDD 262 7/14/2014 7:44:15 PM

FIGURE 21.12 Hepatic veins. There are normally three hepatic veins: right (R ), middle (M ),
and left (L). Note the segments drained. UV is the umbilical vein, which normally drains
part of Sg4 into the left hepatic vein. The latter is proof that the terminal portion of the
left vein lies in the intersectional plane of the left liver. (Copyright Washington University
in St. Louis.)
In about 10% of individuals, there is more than one large right hepatic
vein; in addition to the right superior hepatic vein (normally called the
right hepatic vein), which enters the IVC just below the level of the dia-
phragm, there is a right inferior hepatic vein, which enters the IVC 5 to
6 cm below this level (Fig. 21.12).
The caudate lobe is drained by its own veins—several short veins that
enter the IVC directly from the caudate lobe. When performing a classi-
cal right hepatectomy, caudate veins are divided in the preliminary portion
of the dissection. As dissection moves up the anterior surface of the vena
cava to isolate the right hepatic vein, one encounters a bridge of tissue lat-
eral to the IVC referred to as the “inferior vena cava ligament.” It connects
the posterior portion of the right liver to the caudate lobe behind the IVC.
This bridge of tissue usually consists of fibrous tissue, but occasionally is a
bridge of liver parenchyma.
The Plate/Sheath System of the Liver

There is a system of fibrous plates and sheaths, which lies on the ventral
surface of the liver and extends into it. Knowledge of the system is of impor-
tance in liver surgery. The fundamentals of the system can be understood
by imagining a shirt with the front cut away to leave only the back and the
sleeves (Fig. 21.13 inset). The back of the shirt would be a plate, and the
sleeves would be sheaths. In the true plate/sheath system, there are four
plates (hilar, cystic, umbilical, and arantian) and several sheaths (Fig. 21.13).
0002086387.INDD 263 7/14/2014 7:44:15 PM

FIGURE 21.13 Plate–sheath system of the liver. A. A schematic of a plate with two
sheaths (see text) B. plate sheath system. (From Strasberg SW, Linehan DC, Hawkins WG.
Isolation of right main and right sectional portal pedicles for liver resection without hepa-
totomy or inflow occlusion. J Am Coll Surg 2008;206(2):390–396.)
The hilar plate, the most important plate in liver surgery, is a flat sheet, lying
in the coronal plane, posterior to the main bilovascular structures in the
porta hepatis. However, the upper part curves forward to enclose the right
and left bile ducts, the most superior structures in the porta hepatis. It is
this taut, firm, upper curved edge of the hilar plate that is dissected free
from the underside of the liver when “lowering the hilar plate.”
Coming off the right side of the hilar plate like a sleeve is the sheath
of the right portal pedicle. It extends into the liver surrounding the portal
structures that enter it on its free edge, that is, the portal vein, hepatic
artery, and bile duct (Fig. 21.14). The combined structure consisting of
sheath and contents is the right portal pedicle. As the right portal pedicle
enters the liver, it divides into a right anterior and right posterior portal
pedicle supplying the respective sections and then segmental pedicles
supplying the four segments. On the left side, only the segmental struc-
tures are sheathed.
002086387.INDD 264 7/14/2014 7:44:16 PM

FIGURE 21.14 Isolation of the right portal pedicle and sectional pedicles by technique of
dissection on surface of pedicles. No inflow occlusion or separate hepatotomies are used
(see Fischer et al., 2011). The umbilical tape in the upper right of the photograph is around
the bridge of the liver tissue over the umbilical fissure. (From Strasberg SW, Linehan DC,
Hawkins WG. Isolation of right main and right sectional portal pedicles for liver resection
without hepatotomy or inflow occlusion. J Am Coll Surg 2008;206(2):390–396.)
The cystic plate is the ovoid fibrous sheet on which the gallblad-
der lies (Fig. 21.13). In its posterior extent, the cystic plate narrows to
become a stout cord that attaches to the anterior surface of the sheath
of the right portal pedicle. The latter is a point of anatomical importance
for the surgeon wishing to expose the anterior surface of the right por-
tal pedicle, since this cord must be divided to do so. With severe chronic
inflammation, the cystic plate may become shortened and thickened so
that the distance between the top of the cystic plate and the right portal
pedicle is likewise much shorter than usual. This places the structures in
the right pedicle in danger during cholecystectomy in which dissection
is performed “top down” as a primary strategy. The other plates are the
umbilical and arantian, which underlie the umbilical portion of the left
portal vein and the ligamentum venosum respectively (Fig. 21.13). The
other sheaths carry segmental bilovascular pedicles of the left liver and
caudate lobe.
Liver Capsule and Attachments

The liver is covered with a thin fibrous capsule that covers the entire organ
except for the large “bare area” posteriorly. There, the organ is in contact
with the IVC and diaphragm. The bare area stretches superiorly to include
the termination of the three hepatic veins and ends in a point, which is
also where the attachment of the falciform ligament ends. The limit of the
bare area, where the peritoneum passes between the body wall and the
liver, is called the coronary ligament. It is one of three structures, which
connect the liver to the abdominal wall “dorsally,” the other two being the
right and left triangular ligaments. The liver also has another much smaller
bare area, where the hepatoduodenal ligament and the lesser omentum
attach on the “ventral” surface. It is here that the portal structures enter
the liver at the hilum. The ligamentum teres (“round”) is the obliterated
left umbilical vein and runs in the free edge of the falciform ligament from
the umbilicus to the termination of the umbilical portion of the left portal
vein. The falciform ( falciform = “scythe shaped”) is the filmy fold that runs
in between the anterior abdominal wall above the umbilicus and attaches
to the anterior surface of the liver between the left medial and left lateral
sections. The ligamentum venosum, the residual ductus venosus, is a filmy
0002086387.INDD 265 7/14/2014 7:44:17 PM

cord that extends from the termination of the transverse portion of the
left portal vein to the left hepatic vein along the junction of the left lateral
section and the caudate lobe.
GALLBLADDER AND EXTRAHEPATIC BILE DUCTS

Gallbladder
The gallbladder lies on the cystic plate. The lower edge of the gallbladder
forms one side of the hepatocystic triangle. The other two sides are the
right side of the CHD and the liver. Eponyms covering this anatomy (Calot,
Moosman, etc.) are confusing and should be abandoned. The hepatocystic
triangle contains the cystic artery and cystic node and sometimes a portion
of the right hepatic artery as well as fat and fibrous tissue. Clearance of this
triangle along with the isolation of the cystic duct and elevation of the base
of the gallbladder off the lower portion of the cystic plate gives the “criti-
cal view of safety,” which we have described for identification of the cystic
structures during laparoscopic cholecystectomy.
A large number of minor curiosities of the gallbladder, for example,
phrygian cap, have been described. The following are anomalies of impor-
tance to the biliary surgeon.
Double Gallbladder
This is a rare anomaly but can be the cause of persistent symptoms after
resection of one gallbladder. A gallbladder may also be bifid, which usually
does not cause symptoms, or have an hourglass constriction that may cause
symptoms due to obstruction of the upper segment.
Agenesis of the Gallbladder

Agenesis occurs in about 1/8,000 patients. It can be difficult to recognize.
When agenesis is suspected, it may be confirmed by axial imaging. If doubt
remains, laparoscopy is definitive.
Cystic Duct
A tubular structure normally 1 to 2 cm in length and 2 to 3 mm in diameter.
It usually joins the CHD at an acute angle to form the CBD. The cystic duct
normally joins the CHD approximately 4 cm above the duodenum. However,
the cystic duct may enter at any level up to the right hepatic duct and down
to the ampulla. The cystic duct may also join the right hepatic duct either
when the right duct is in its normal position or in an aberrant location. The
cystic duct contains a spiral valve.
There are three patterns of confluence of the cystic duct and CHD
(Fig. 21.15). In 20% of patients in which there is a parallel union, injury to the
CHD may occur by dissecting low on the cystic duct (Fig. 21.15). When mak-
ing a choledochotomy at this level, the incision should be started slightly
to the left side of the midplane of the bile duct in order to avoid entering a
septum between the fused cystic duct/common hepatic duct. When per-
forming cholecystectomy, the cystic duct should be occluded in such a way
that there is a visible section of cystic duct below the clip closest to the CBD.
Although a gallbladder with two cystic ducts has been described, it is
an extreme rarity. When two “cystic ducts” are identified, it is likely that cys-
tic duct is congenitally short or has been effaced by a stone and that the two
structures thought to be dual cystic ducts are, in fact, the CBD and the CHD.
Cystic Artery
The cystic artery is usually 1 to 2 mm in diameter and normally arises from
the right hepatic artery in the hepatocystic triangle. The cystic artery may
002086387.INDD 266 7/14/2014 7:44:17 PM

FIGURE 21.15 The three types of cystic duct/common hepatic duct confluence. The parallel
union confluence is shown in the middle. Dissection of this type of cystic duct (arrow) may
lead to injury to the side of CHD. During laparoscopic cholecystectomy, this is often a cau-
tery injury. (Adapted from Warrren KW, McDonald WM, Kune GA. Bile duct strictures: new
concepts in the management of an old problem. In: Irvine WT (ed). Modern trends in surgery.
London, UK: Butterworth, 1966. Copyright Washington University in St. Louis.)
arise from a right hepatic artery that runs anterior to the CHD. The cystic
artery may also arise from the right hepatic artery on the left side of the CHD
and run anterior to this duct, while the right hepatic artery runs behind it.
Such cystic arteries tend to tether the gallbladder and make dissection of
the hepatocystic triangle more difficult. The cystic artery may arise from a
replaced right hepatic artery arising from the SMA. In this case, the cystic
artery (and not the cystic duct) tends to be in the free edge of the fold lead-
ing from the hepatoduodenal ligament to the gallbladder. This should be
suspected whenever the “cystic duct” looks smaller than the “cystic artery.”
Usually, the cystic artery runs for 1 to 2 cm to meet the gallbladder
superior to the insertion of the cystic duct. The artery ramifies into an ante-
rior and posterior branch at the point of contact with the gallbladder. These
branches continue to divide on their respective surfaces. Sometimes, the
cystic artery divides into branches before the gallbladder edge is reached.
In that case, the anterior branch may be mistaken for the cystic artery
proper and the posterior branch may not be discovered until later in the
dissection—sometimes by inadvertent division with hemorrhage. The
artery may ramify into several branches before arriving at the gallbladder
giving the impression that there is no cystic artery. The anterior and poste-
rior branches may arise independently from the right hepatic artery, giving
rise to two distinct cystic arteries. There are many other variations.
Multiple small cystic veins drain into intrahepatic portal vein branches
by passing into the liver around or through the cystic plate. Sometimes,
there are cystic veins in the hepatocystic triangle that run parallel to the
cystic artery to enter the main portal vein.
Cystic Plate
The cystic plate has been described previously. Small bile ducts may pen-
etrate the cystic plate to enter the gallbladder. These “ducts of Luschka” are
very small, usually submillimeter accessory ducts. However, when divided
0002086387.INDD 267 7/14/2014 7:44:18 PM

during a cholecystectomy, postoperative bilomas may occur. Bilomas and

hemorrhage may also be caused by penetration of the cystic plate during
dissection. In about 10% of patients, there is a large peripheral bile duct
immediately deep to the plate, disruption of which will cause copious bile
drainage. The origin of the middle hepatic vein is also in this location,
and if it is injured, massive hemorrhage may ensue. There is areolar tissue
between the muscularis of the gallbladder and the cystic plate. At the top
of the gallbladder, the layer is very thin. As one progresses downward, the
areolar layer thickens. If dissection from the top of the gallbladder down-
ward is carried on the gallbladder leaving the areolar tissue on the cystic
plate, one will arrive onto the posterior surface of the cystic artery and
cystic duct. Conversely, if dissection is carried downward on the cystic
plate leaving the areolar tissue on the gallbladder, one will arrive onto the
surface of the right portal pedicle. If this is not anticipated, structures in
the right portal pedicle may be injured. Therefore, the proper plane of dis-
section is between the gallbladder and the areolar tissue.
Extrahepatic Bile Ducts

The CHD is formed by the union of the right and left hepatic ducts. The
union normally occurs at the right extremity of the base of Sg4 anterior and
superior to the bifurcation of the portal vein. The CHD travels in the right
edge of the hepatoduodenal ligament for 2 to 3 cm where it joins with the
cystic duct to form the CBD. The latter has a supraduodenal course of 3 to
4 cm and then passes behind the duodenum to run in or occasionally behind
the pancreas to enter the second portion of the duodenum. The external
diameter of the CBD varies from 5 to 13 mm when distended to physiologic
pressures. However, the duct diameter at surgery, that is, in fasting patients
with low duct pressures, may be as small as 3 mm. Radiologically, the inter-
nal duct diameter is measured on fasting patients. Under these conditions,
the upper limit of normal is about 8 mm. Size should never be used as a
sole criterion for identifying a bile duct. Caution is required in situations
in which a structure seems larger than the expected norm. Although the
cystic duct may be enlarged due to passage of stones, a surgeon should take
extra precautions before dividing a “cystic duct” that is greater than 2 mm
in diameter because the CBD can be 3 mm in diameter and aberrant ducts
may be smaller.
Anomalies of Extrahepatic Bile Ducts

As already noted, there are biliary anomalies of the right and left ductal
systems, which can affect outcome of hepatic surgery. The same is true for
biliary surgery. The most important clinical anomaly is low insertion of the
right hepatic ducts referred to above. Because of its low location, it may be
mistaken to be the cystic duct and injured during cholecystectomy. This is
even more likely to occur when the cystic duct unites with an aberrant duct
as opposed to joining the CHD. The anomalies of a right hepatic duct enter-
ing the gallbladder and the absent CHD have been discussed above. Left
hepatic ducts can also join the CHD at a low level. They are less prone to be
injured since the dissection during cholecystectomy is on the right side of
the biliary tree.
Extrahepatic Arteries
The course of these arteries has been described above. Anomalies of the
hepatic artery may be important in gallbladder surgery. Normally, the right
hepatic artery passes posterior to the bile duct (80%) and gives off the cys-
tic artery in the hepatocystic triangle. However, in 20% of cases, the right
hepatic artery runs anterior to the bile duct. The right hepatic artery may
002086387.INDD 268 7/14/2014 7:44:18 PM

lie very close to the gallbladder, and chronic inflammation can draw the
right hepatic artery directly onto the gallbladder, where it lies in an inverse
U-loop and is prone to injury. In the “classical injury” in laparoscopic chole-
cystectomy in which the CBD is mistaken for the cystic duct, an associated
right hepatic artery injury is very common, since that right hepatic artery
is considered to be the cystic artery.
Blood Supply of Bile Ducts

Many studies, dating back to the 19th century, have examined the blood
supply of the extrahepatic bile ducts in cadaveric specimens. An important
observation made by Rappoport is that the bile ducts are supplied solely by
the hepatic artery, unlike the liver that has a dual blood supply from the artery
and the portal vein. The arterial blood supply has three anatomic elements—
afferent arteries, marginal arteries, and epicholedochal plexus (Fig. 21.16).
The afferent vessels are branches of the hepatic arteries and less com-
monly of the SMA or other upper abdominal arteries. The most constant
and important artery supplying the bile duct is the posterior superior
pancreaticoduodenal artery, usually the first branch of the gastroduode-
nal artery. Arterial twigs are sent to the duct as the artery winds around
the lower end of the duct. These branches supply the retroduodenal and
intrapancreatic bile duct but also ascend the bile duct to give supply to the
supraduodenal bile duct. The lowest portion of the duct near the ampulla
is also supplied by the anterior superior pancreatic artery and the inferior
pancreaticoduodenal artery. Other vessels, which commonly send afferents
to the supraduodenal bile duct, are the proper hepatic artery, cystic artery,
Epicholedochal plexus Marginal arteries
Right hepatic artery

Left hepatic artery
Cystic artery
Proper hepatic artery
Posterior superior Common hepatic artery

pancreato-duodenal
artery
Gastroduodenal artery
FIGURE 21.16 Blood supply to bile ducts. The three elements—the

afferent arteries, marginal arteries, and epicholedochal plexus—
are shown. (From Fischer JE, Jones DB, Pomposelli FB, et al.
Fischer’s mastery of surgery, 6th ed. Philadelphia, PA: Lippincott
Williams & Wilkins, 2011.)
0002086387.INDD 269 7/14/2014 7:44:30 PM

and artery to segment 4. However, virtually all extrahepatic arteries arising

from the common hepatic artery can send small branches to the bile ducts.
Also, body wall collaterals such as phrenic arteries can at times supply the
bile ducts. The idea that the extrahepatic bile duct, that is, CHD and CBD, is
supplied by the arteries that join it only at the bottom and top of its course
is incorrect. Supplying arteries from the cystic artery, right and left hepatic
arteries, and proper hepatic artery may join it along its length.
The afferent vessels empty into the longitudinal or “marginal” arter-
ies, which run parallel to the long axis of the bile duct. These vessels are
disposed at 3, 9, and, less commonly, 12 o’clock on the CBD/CHD or run
across the top of the confluence of the right and left hepatic ducts. This hilar
marginal artery is of great importance in maintaining blood supply to the
liver when one hepatic artery (right or left) is occluded.
The third element, the “epicholedochal plexus,” is a fine arterial plexus
that lies on and surrounds the entire CBD and the left and right hepatic
ducts. The part around the hepatic ducts is the hilar component of the epi-
choledochal plexus. The vessels of the plexus tend to run along the long
axis of the ducts so that on the common duct many of the vessels are verti-
cal while that around the confluence and the right and left ducts they are
disposed horizontally. Branches of the epicholedochal plexus pierce the
bile duct and supply it through the deeper and finer intracholedochal plexi
within the wall of the bile duct.
Transection of the bile duct may result in ischemia of the duct. For
instance, if the duct is transected at the level of the duodenum, ischemia of
a portion of the bile duct above this level may occur since blood flow origi-
nating from the superior pancreaticoduodenal artery and passing up along
the marginal artery is cut off. This problem is thought to be an important
contributory cause to the frequent failure of choledochocholedochostomy
as a form of biliary reconstruction. To avoid this problem, hepaticojeju-
nostomy is used and the bile duct is trimmed back to within 1 cm of the
confluence.
Suggested Readings
Couinaud C. Le Foie. Etudes Anatomiqucs et Churgicales. Paris, France: Masson & Cie,
1957.
Fischer JE, Jones DB, Pomposelli FB, et al. Fischer’s mastery of surgery, 6th ed.
Philadelphia, PA: Lippincott Williams & Wilkins, 2011.
Goldsmith NA, Woodburne RT. The surgical anatomy pertaining to liver resection.
Surg Gynecol Obstet 1957;195:310–318.
Gunji H, Cho A, Tohma T, et al. The blood supply of the hilar bile duct and its rela-
tionship to the communicating arcade located between the right and left hepatic
arteries. Am J Surg 2006;192:276–280.
Healey JE, Schroy PC. Anatomy of the biliary ducts within the human liver; Analysis of
the prevailing pattern of branchings and the major variations of the biliary ducts.
Arch Surg 1953;66:599–616.
Michels NA. Blood supply and anatomy of the upper abdominal organs. Philadelphia
and Montreal: J.B. Lippincott, 1955.
Northover JMA, Terblanche J. A new look at the arterial supply of the bile duct in man
and its surgical implications. Brit J Surg 1979;66:379–384.
Strasberg SM, Linehan DC, Hawkins WG. Isolation of right main and right sectional
portal pedicles for liver resection without hepatotomy or inflow occlusion. J Am
Coll Surg 2008;206:390–396.
Strasberg SM, Phillip C. Use and dissemination of the Brisbane 2000 nomenclature of
liver anatomy and resections. Ann Surg 2013;257:377–382.
Terminology Committee of the IHPBA. The Brisbane 2000 terminology of liver anat-
omy and resections. HPB 2000;2:333–339.
002086387.INDD 270 7/14/2014 7:44:30 PM

22 Infections in Biliary Surgery
Massimo Arcerito and Rebecca M. Minter
INTRODUCTION
Although infection is not often the primary cause of disease in the biliary
system, it is a common complication. Obstructing conditions such as gall-
stones, benign strictures, or endoluminal or extraluminal bile duct malig-
nancies lead susceptible patients to develop infectious complications,
which are caused by microorganisms. Gram-negative aerobic organisms
(Enterobacteriaceae species), anaerobes, fungi, or a combination colonize
the biliary system and in settings of obstruction can cause cholangitis.
Biliary infections, also called infectious cholangitides, can manifest acutely
or more indolently. Local infections can lead to cholangitis and subsequent
liver abscesses that may spread hematogenously to cause septicemia or sep-
tic shock, requiring immediate endoscopic or surgical intervention. Bacteria
predominantly cause infectious cholangitides in the Western society, while
in other parts of the world parasites play a larger role in the development of
biliary tract infections. In the last 30 years, immunocompromised patients
(e.g., patients with AIDS or posttransplant) with viral cholangitides repre-
sent a more difficult population to treat by clinicians and surgeons.
It is well recognized that patients undergoing hepato-pancreatico-
biliary surgery are at risk for infectious complications that are not limited
to skin or intra-abdominal locations, but also within the biliary system and
the liver parenchyma, potentially leading to the development of pyogenic
liver abscesses. A very fine balance exists between the host defense mecha-
nisms and infections of the biliary system when obstructing lesions of the
biliary tree are addressed by endoscopic, percutaneous, or surgical means.
BACKGROUND
The biliary system and the liver represent a unique environment to protect
the human body from hematogenous dissemination of toxins, microorgan-
isms, and enteric pathogens delivered by the gastrointestinal system to the
liver. The inferior and superior mesenteric venous systems drain into the
portal vein, ultimately reaching the hepatic sinusoids where the blood is
detoxified before reaching the suprahepatic veins. Typically, bile within the
gallbladder and the biliary system is sterile in the absence of gallstones,
obstruction, or communication with the gastrointestinal tract via either
disease-driven fistula or endoscopic or surgical interventions. It is the
presence of infected bile, or bacteribilia—secondary to gallstones, benign
and malignant obstruction, diagnostic or therapeutic interventions—that
cause infections in the hepatobiliary system.
BACTERIOLOGY
In patients with symptomatic gallstones or acute or chronic cholecysti-
tis, the incidence of positive bile cultures ranges between 11% and 30%.
271
0002105272.INDD 271 7/14/2014 7:47:29 PM

Acute cholecystitis, often characterized by fever and leukocytosis,

demonstrates a higher rate of positive bile cultures as compared with
chronic cholecystitis (46% vs. 22%). Cholangitis secondary to endobiliary
stents, stones, or tumor presents with a positive bile culture in 100% of cases.
The most common bacteria isolated are of gastrointestinal origin, specifi-
cally Escherichia coli (25% to 50%) followed by Klebsiella (15% to 20%) and
Enterobacter species (5% to 10%) are the most common Gram-negative spe-
cies. Anaerobes such as Clostridium perfringens and Bacteroides can also be
present but are more commonly observed in the setting of recurrent chol-
angitis or following hepatobiliary surgical interventions. Candida species
have been more often identified in the bile of critically ill patients. Achieving
adequate drainage of the biliary tree by treating the underlying obstruction
is critical for successful treatment, typically in conjunction with broad-spec-
trum antibiotic therapy covering both aerobic and anaerobic pathogens.
HOST DEFENSE MECHANISM

Physical, chemical, and immunologic properties are required to protect the
liver from invasion by pathogens and toxins. Table 22.1 represents the fea-
tures of the hepatobiliary host defense mechanism that function in a coor-
dinated fashion to prevent invasion by microorganisms within the biliary
system.
Division of the bile canaliculi from the bloodstream is required to
protect the biliary system from pathogen invasion by the vascular system,
and vice versa. This extraordinary property is provided by the hepatic tight
junctions. These junctions are able to build a barrier between the biliary
channels and hepatic sinusoids. The liver is the core metabolic center in
humans and is responsible for fundamental physiologic functions includ-
ing fatty acid, amino acid, and carbohydrate metabolism; bile secretion;
and detoxification. Because of the highly vascularized nature of the liver
and its hepatocytes, a microvascular network is maintained in place by
the hepatic tight junctions. Deregulation of tight junction expression dis-
mantles the parenchymal network, causing liver disease and malignant
transformation. Hepatic tight junctions also represent a physical barrier
within the liver preventing ascending biliary infection. Any obstruction of
TABLE
22.1 Hepatobiliary Host Defense Mechanisms
Defense Mechanisms
Physical Antegrade flow of bile into the duodenum
Sphincter of Oddi preventing reflux of gastrointestinal
content into biliary system
Tight junctions between bile canaliculi and hepatic
sinusoids
Chemical Bile salts possess bacteriostatic and bacteriocidal
properties
Immunologic Kupffer cells phagocytose bacteria and bacterial
products and play a critical role in bilirubin
metabolism through phagocytosis of senescent
erythrocytes
IgA secreted by biliary epithelium
Fibronectin and complement (C3, C4, and Factor B)
play a critical role in opsonizing pathogens in liver
002105272.INDD 272 7/14/2014 7:47:29 PM

Chapter 22 / Infections in Biliary Surgery 273
the portal vein, pre-, intra- or posthepatic, will cause portal hypertension
with the development of portosystemic shunts. Those shunts will bypass
the liver parenchyma and will allow the toxins and enteric pathogens to
reach the systemic circulation.
Additional physical features of the hepatobiliary system serve to pre-
vent ascending infection. Specifically, the sphincter of Oddi that prevents
reflux of enteric content into the biliary tree when bile flows antegrade into
the duodenum represents a dynamic physical barrier against ascending
biliary infection. After cholecystectomy, there may be a decreased force of
antegrade bile flow due to the loss of gallbladder contraction with retro-
grade reflux of enteric contents leading to colonization of the biliary tree.
In addition, any surgical or endoscopic intervention on the biliopancreatic
sphincter (e.g., sphincterotomy, placement of endobiliary stent, or creation
of a biliary–enteric bypass) leads to loss of this physical barrier, permitting
direct reflux of enteric content into the biliary tree. The mere presence of
bacteria in the bile does not always lead to infection; however, in the setting
of distal obstruction one must watch closely for signs and symptoms of an
ascending infection.
The second protective element in the host defense mechanisms is
the chemical component. Specifically, the presence of bile salts in the bile
provides an excellent protection against enteric pathogens possessing both
bacteriostatic and bactericidal properties. Jaundice due to mechanical
obstruction of the bile duct leads to overgrowth of intestinal flora due to
the absence of bile and bile salts within the gastrointestinal tract, predis-
posing patients to develop cholangitis when the obstruction is treated with
endobiliary or surgical decompression. Fungal contamination of the bile is
commonly observed in patients with biliary obstruction. This fungal coloni-
zation is secondary to the absence of bile salts that also possess antifungal
properties, particularly against Candida albicans species.
The third protective element is the immunologic and humoral com-
ponent. The liver parenchyma contains several immunologic active cells
whose function is to alter and metabolize toxins coming from the gas-
trointestinal tract. Kupffer cells represent the largest population of tis-
sue macrophages. Kupffer cells possess Fc, C3, and scavenger receptors
that are known to phagocytize a wide variety of both opsonized and non-
opsonized materials including bacteria and bacterial products. In addi-
tion, Kupffer cells play a defined role in bilirubin metabolism, controlling
infections in the biliary system. Approximately 75% of bilirubin is derived
from the breakdown of senescent erythrocytes by macrophages. Two iso-
enzymes HO-1 and HO-2 are involved in this process. HO-1, also known
as heat shock protein-32, is induced by stressors and is contained in the
endoplasmic reticulum and perinuclear envelope of Kupffer cells in the
liver. Overall, Kupffer cells are responsible for the removal of senescent
erythrocytes from the blood circulation by scavenger receptors. Depletion
of Kupffer cells in the liver reduces HO-1 expression and bilirubin produc-
tion, predisposing to a variety of biliary infections. Liver resections also
reduce the presence of Kupffer cells and therefore reduce their immuno-
logic effects in the liver initially following resection. Immunoglobulin A
(IgA), as part of the humoral component, is produced and secreted into
the hepatic system by the gallbladder and intra- and extrahepatic biliary
epithelium protecting against infectious pathogens. In addition, comple-
ment C3 and C4, factor B, and fibronectin play a major role in opsonizing
pathogens and helping the biliary system to defend itself from the con-
tinuous exposure to pathogens. Any alteration of these defense mecha-
nisms predisposes to colonization and development of infection in the
hepatobiliary system (Table 22.1).
0002105272.INDD 273 7/14/2014 7:47:29 PM

BILIARY OBSTRUCTION
The etiology of biliary obstruction can be secondary to intrinsic causes like
cholangiocarcinoma, choledocholithiasis, or ampullary stenosis, extrinsic
causes like periampullary or pancreatic neoplasia or pancreatitis involving
the distal portion of the bile duct, or functional causes such as sphincter
of Oddi dysfunction. The use of endobiliary stents can often resolve the
obstruction, but predisposes the patient to develop biliary infection due
to the communication this creates with the gastrointestinal tract. In the
presence of biliary obstruction, the lack of bile flow is considered to be
a primary cause of bacterial overgrowth in the biliary system. Likewise,
obstruction leads to biliary hypertension with subsequent breaking down
of the hepatocyte tight junction and development of cholangiovenous
reflux. This phenomenon allows the gastrointestinal endotoxins to colo-
nize the biliary–portal system and can lead to the formation of pyogenic
liver abscesses or biliary sepsis. The absence of bile salts with their bacte-
riostatic and bacteriocidal properties, which occurs in the setting of biliary
obstruction, leads to an increase in the number of enteric pathogens in the
intestinal tract.
No one clearly delineated mechanism defines how biliary obstruction
specifically predisposes one to infectious complications. One theory sug-
gests that Kupffer cell metabolism is diminished in the presence of biliary
obstruction. This leads to bacterial overgrowth with inability of the exist-
ing Kupffer cells to clear the bacteria and bacterial products adequately.
Bacteria can then translocate to the systemic circulation through the
hepatic–sinusoid tight junctions. In addition, IgA production is limited by
biliary obstruction. Clinically, the presence of bacteria in the biliary tree
does not translate into cholangitis, but the combination of biliary obstruc-
tion and increased pressure in the biliary system determines a cascade of
events that can often lead to acute cholangitis.
INFECTIONS IN BILIARY SURGERY

Human bile is typically sterile but may become infected in the presence of
gallbladder and common bile duct stones or intrinsic or extrinsic inflam-
matory or neoplastic diseases leading to obstruction. The pathophysiology
of infection is due either to translocation of intestinal pathogens via the por-
tal system or more commonly due to biliary intervention with incomplete
drainage. Any invasive endoscopic, radiologic, or surgical procedure on the
hepatobiliary system interferes with the usual host defense mechanism,
and the introduction of an indwelling biliary stent through any means con-
stitutes a foreign body within the biliary system. Biliary–enteric anastomo-
sis, sphincterotomy and sphincteroplasty, disruption of blood supply either
locally to the liver and biliary tree or systemically (shock), blood transfu-
sions, and the immune status of the patient are all additional important
contributors to the development of an ascending biliary infection.
Hepatic resection can also predispose to the development of bili-
ary infections. It is mandatory to analyze the hepatic reserve in patients
undergoing resection of hepatic parenchyma, particularly in those who
will require a bilioenteric anastomosis such as in the setting of cholan-
giocarcinoma. The reticuloendothelial system immunologic function is
reduced after hepatic resection secondary to loss of functional liver mass.
Increasingly, preoperative portal vein embolization is being utilized in prep-
aration of planned extended hepatic resection to provide sufficient post-
operative hepatic metabolic and host defense function by inducing growth
of the planned hepatic remnant. A recent Cochrane review of 521 patients
evaluating the use of perioperative antibiotics, duration of antibiotics, and
002105272.INDD 274 7/14/2014 7:47:29 PM

use of probiotics did not identify any clear predictors of morbidity or mor-
tality related to postoperative infectious complications for these variables.
ACUTE ASCENDING CHOLANGITIS

Cholangitis is caused by an acute bacterial infection of the bile duct. In
1877, Charcot first described the triad of fever, jaundice, and right upper
quadrant pain. When shock and altered mental status are included in the
classic clinical scenario, it is referred to as Reynolds pentad. The develop-
ment of cholangitis requires biliary bacterial contamination, stagnant bile,
and increased intrabiliary pressure (>20 cm H2O). Cholangitis is a serious
medical problem; mortality ranges between 3.5% and 65%. Interestingly, the
severity of acute cholangitis has been shown to correlate with intrabiliary
pressure. The most common cause of cholangitis is partial biliary obstruc-
tion. Complete obstruction is less likely to be associated with bacterobilia
and therefore is less likely to present as cholangitis. Elevated biliary pres-
sure increases biliary ductule permeability, allowing pathogens to enter the
bloodstream. Two decades ago, the most common cause of acute cholan-
gitis was choledocholithiasis. In more recent years, other causes of chol-
angitis have increased, including altered biliary anatomy due to surgical
intervention and instrumentation of the biliary tree. Malignancy accounts
for 10% to 30% of all cases of cholangitis. Bile cultures are polymicrobial in
30% to 80% of patients who have undergone placement of endobiliary stents
or biliary–enteric anastomosis; these patients often are colonized with anti-
biotic-resistant organisms. The most common aerobic isolated organisms
in culture from these patients are E. coli, Klebsiella, Enterococcus species;
Enterobacter, Pseudomonas, Serratia and Proteus species, and Bacteroides
fragilis are the most common anaerobic enteric pathogens. C. albicans is
the most common fungus in bile cultures in patients who have undergone
biliary manipulation.
The diagnosis of acute cholangitis is made based upon clinical pre-
sentation. All patients must be aggressively resuscitated and started on
broad-spectrum antibiotics immediately. Blood cultures are obtained at
presentation, and bile cultures are also obtained as soon as possible to
guide antibiotic therapy. Initial antibiotic coverage is broad and is tailored
to the severity of the clinical presentation, the patient’s history of prior anti-
biotic usage, hepatic and renal function, allergies, and any prior culture
data. Typically, community-acquired cholangitis is more commonly asso-
ciated with a single biliary organism (most commonly Klebsiella, E. coli, or
Enterococcus), and treatment with a short course (3 to 5 days) of a penicillin/
β-lactamase inhibitor combination (e.g., piperacillin/tazobactam or ampi-
cillin/sulbactam) is usually effective. Patients presenting with greater
clinical severity often have anaerobic bacteria contributing to the infection,
and patients in the hospital may have infection due to resistant and/or mul-
tiple organisms such as Pseudomonas, methicillin-resistant Staphylococcus
aureus, and vancomycin-resistant enterococci. Antibiotic choice must be
tailored to address these conditions. Patients presenting with moderate
or severe cholangitis should also be treated with intravenous penicillin/
β-lactamase inhibitors; however, if resistant organisms are suspected then
the addition of appropriate antibiotics for methicillin-resistant bacteria like
vancomycin or daptomycin for coverage of suspected vancomycin-resistant
Enterococcus should be added. Double coverage for Pseudomonas should
also be considered if the infection is occurring in a hospitalized patient
who is not responding to first-line therapy. If the patient does not respond
to the initial antibiotic choice, then addition of fluoroquinolones and car-
bapenems should be considered along with antifungal coverage. Antibiotics
0002105272.INDD 275 7/14/2014 7:47:29 PM

TABLE
22.2 Antibiotic Choice for Treatment of Acute Cholangitis
Duration of
Clinical Severity Antibiotic Selection Treatment
Mild (Community Pipericillin/tazobactam or ampicillin/ 3–5 d
Acquired) sulbactam
Moderate/Severea Pipericillin/tazobactam and Minimum 7 db
metronidazole (first-line)
Addition of carbapenems or
fluoroquinolones, and Fluconazole
if fail to respond clinically
Hospital Acquired/ Pipericillin/tazobactam and Minimum 7 db
Resistant metronidazole (first-line)
Organisms Addition of carbapenems or
Anticipateda fluoroquinolones, and Fluconazole
if fail to respond clinically
Addition of Vancomycin for methicillin-
resistant S. aureus, daptomycin for
vancomycin-resistant Enterococcus,
or additional anti-Pseudomonas
coverage should also be considered
a
Antibiotic regimen should be narrowed once culture data are available.
b
Course will be determined based upon clinical context and response to therapy.
should be continued a minimum of 7 days for moderate to severe cases of

cholangitis; however, the clinical context, culture data, and response of the
patient will ultimately dictate the choice and length of antibiotic course
required (Table 22.2).
Drainage of the biliary tree is a central tenet to the treatment of
ascending cholangitis, and must occur as soon as possible in a patient with
sepsis secondary to ascending cholangitis. Thus, following administration
of antibiotics and initiation of resuscitation and necessary supportive care,
either endoscopic retrograde cholangiopancreatography (ERCP) or percu-
taneous transhepatic cholangiography (PTC) is the next step in manage-
ment. If the patient presents with hemodynamic instability or respiratory
compromise, then appropriate supportive management must be initiated
prior to transporting the patient for a planned biliary drainage procedure.
Endoscopic decompression is preferred; however, PTC drainage may be
necessary if endoscopic decompression of the biliary tree is not possible. If
a patient presents with a clinical picture of ascending cholangitis and has
an indwelling biliary stent, then this stent and the biliary tree should be
interrogated to ensure that there is complete drainage of the biliary system
and that the stent is functioning and is not occluded. Plastic indwelling bili-
ary stents require replacement approximately every 8 to 12 weeks (depend-
ing on stent diameter and number of stents) to prevent occlusion. Metal
biliary stents (typically only placed in patients with biliary obstruction sec-
ondary to advanced malignancy) are larger in diameter and have a lower
incidence of occlusion; however, sludge and stones can develop within and
above these stents leading to the development of cholangitis. Endoscopic
and percutaneous intervention is then required to achieve drainage of the
biliary system. If the underlying cause of the cholangitis is choledocholi-
thiasis, then clearance of the biliary system can also typically be accom-
plished with either endoscopic or percutaneous interventions to clear the
002105272.INDD 276 7/14/2014 7:47:29 PM

common bile duct, or if a stone is impacted to drain the biliary tree either
internally or externally with a biliary stent. Operative intervention is rarely
the first-line therapy for biliary drainage as endoscopic and percutaneous
drainage have lower morbidity and mortality and can usually be achieved
in a timely manner. If, however, these modalities of biliary drainage are not
available for a decompensating patient with severe ascending cholangitis,
then operative decompression of the biliary tree should be pursued.
ACUTE CHOLECYSTITIS
Acute cholecystitis represents sudden inflammation of the gallbladder, and
is typically associated with severe right upper quadrant pain, fever, and leu-
kocytosis. Usually, this disease is secondary to cholelithiasis with the acute
development of an impacted stone at Hartmann’s pouch or the cystic duct
in the majority of patients (90% to 95%). Acalculous cholecystitis presents
in a similar manner to acute cholecystitis without gallstone involvement
and usually occurs in critically ill patients representing approximately 10%
of acute cholecystitis cases. Infected bile might play a role in the develop-
ment of acute cholecystitis; however, many patients with acute cholecysti-
tis do not have infected bile. Csendes et al. analyzed 467 patients with bile
samples obtained for aerobic and anaerobic culture from healthy control
patients and patients with various biliary diseases including acute chole-
cystitis, cholelithiasis, and gallbladder hydrops. Only 22% to 46% of patients
with the included biliary diseases had a bile culture that was positive for
aerobic or anaeroboic organisms. Healthy control patients had sterile bile
as would be expected. The most common isolated organisms in the bile were
E. coli, Enterococcus species, Klebsiella, and Enterobacter. Complications of
acute cholecystitis include gallbladder empyema, perforation, and emphy-
sematous cholecystitis. Clinically, any prolonged right upper quadrant pain
associated with leukocytosis with or without fever should trigger an evalu-
ation for cholecystitis. Abdominal ultrasound and radionuclide scintigra-
phy represent the two primary diagnostic tests to confirm the diagnosis.
Ultrasound will show the presence of gallbladder wall thickening usually
in the presence of gallstones with pericholecystic fluid, while radionuclide
study will demonstrate uptake of the radiotracer by the liver with excretion
into the common bile duct and duodenum within 1 to 2 hours without filling
or visualization of the gallbladder. Failure to visualize the gallbladder with
radionuclide tracer simply confirms obstruction of the cystic duct but does
not confirm a diagnosis of acute cholecystitis. However, it can be useful for
differentiating symptomatic cholelithiasis from cholecystitis in the setting
of operative decision making.
If a diagnosis of acute cholecystitis is suspected, the role of antibiotic
therapy is debated. Antibiotic therapy is not recommended beyond rou-
tine perioperative antibiotics for uncomplicated acute cholecystitis. The
Infectious Disease Society of America currently recommends that antibiot-
ics should be administered for acute cholecystitis in the setting of a leuko-
cytosis exceeding 12,500 WBC/mm3, fever greater than 38.5 C, suggestion
of emphysematous cholecystitis manifest as evidence of air in the wall of
the gallbladder, and in elderly, diabetic, or immunocompromised patients.
These patients should receive intravenous antibiotics with piperacillin/
tazobactam or a second- or third-generation cephalosporin with duration
tailored to clinical improvement.
Laparoscopic cholecystectomy is typically performed as defini-
tive management of acute cholecystitis unless the patient is felt to have
prohibitive operative risk. In these high-risk patients, percutaneous cho-
lecystostomy can be performed in radiology under ultrasound guidance
0002105272.INDD 277 7/14/2014 7:47:30 PM

and local anesthesia for management of their cholecystitis. Patients with

uncomplicated acute cholecystitis who do not undergo definitive opera-
tive management will usually see resolution of their symptoms 5 to 7 days
after presentation. Cholecystectomy is typically delayed in these settings
to 6 weeks or greater following their initial presentation; however, this
delay in intervention is associated with an increased conversion rate to
an open procedure of approximately 15% to 20% in most recent series.
Acute cholecystitis can present with severe complications including
gallbladder empyema, emphysematous cholecystitis, and perforation of the
gallbladder. Gallbladder perforation is a rare occurrence (5%) in the set-
ting of acute cholecystitis and can occur within a few days or few weeks
of a patient’s presentation with acute cholecystitis. Presenting symptoms of
these serious complications are more severe than the usual presentation of
acute cholecystitis, and prompt surgical intervention is mandatory to pre-
vent significant morbidity or mortality.
CONCLUSIONS
Physical, chemical, and immunologic properties of the liver and biliary
system characterize the host defense mechanisms protecting the hepa-
tobiliary system from infectious organisms. Although infection is not
often the primary cause of biliary tract disease, it is a common complica-
tion. Many patients with gallstones obstructing the biliary tree develop
infectious complications caused by normal gastrointestinal flora such as
Enterobacteriaceae, Klebsiella, E. coli, and B. fragilis. Local infection can
result in cholangitis and pyogenic liver abscesses or invade the bloodstream
causing bacteremia, septicemia or septic shock. Removing the underlying
obstruction in the biliary tree is a prerequisite to successful therapy. Broad-
spectrum antibiotics, covering both aerobes and anaerobes, are almost
universally utilized. Acute cholecystitis and acute cholangitis can be a pri-
mary infection due to gallstones or could be a secondary event due to endo-
scopic, radiologic, or surgical instrumentation to the biliary tree. Treatment
of systemic infection, resuscitation, and treatment of the underlying source
of infection with either cholecystectomy or gallbladder drainage in the set-
ting of acute cholecystitis or relief of biliary obstruction in the setting of
acute cholangitis are critical to the successful treatment of patients pre-
senting with these conditions.
Suggested Readings
Csendes A, Burdiles P, Maluenda F, et al. Simultaneous bacteriologic assessment of
bile from gallbladder and common bile duct in control subjects and patients with
gallstones and common duct stones. Arch Surg 1996;131(4):389–394.
Gurusamy KS, Naik P, Davidson BR. Methods of decreasing infection to improve out-
comes after liver resections. Cochrane Database Syst Rev 2011;(11):CD006933. doi:
10.1002/14651858.CD006933.pub2.
IDSA. ISDA Guidelines. http://www.idsociety.org/idsa_practice_guidelines. 2013.
Kinney KP. Management of ascending cholangitis. Gastrointest Endosc Clin N Am
2007;17(2):289–306. Accessed date: April 1, 2014.
Lillemoe KD. Surgical treatment of biliary tract infections. Am Surg 2000;66(2):138–144.
Mazeh H, Mizrahi I, Dior U, et al. Role of antibiotic therapy in mild acute calculus
cholecystitis: a prospective randomized controlled trial. World J Surg 2012;36(8):
1750–1759.
002105272.INDD 278 7/14/2014 7:47:30 PM

23 Gallbladder Disease
Henry A. Pitt
INTRODUCTION
Gallbladder disease is a major health problem throughout the world. In
the United States, approximately 12% of the population or more than
30 million Americans have gallstones. Currently, in the United States, more
than 750,000 cholecystectomies are performed annually, and the cost of
caring for these patients is estimated to be between 7 and 10 million dollars
per year. To properly manage these patients, surgeons should understand
biliary physiology; the pathogenesis, incidence, risk factors, and natural
history of gallstones; cholecystitis; and biliary dyskinesia. This chapter
also reviews the diagnostic imaging and management options, as well as
expected outcomes.
ETIOLOGY/PATHOGENESIS
Gallstones
Gallstones represent a failure to maintain biliary solutes, primarily choles-
terol and calcium salts, in a liquid state. Gallstones are classified by their
cholesterol content as either cholesterol or pigment stones. Pigment stones
are further classified as either black or brown. Pure cholesterol gallstones are
uncommon (10%) with most cholesterol stones containing calcium salts in
their center, or nidus. In most Western populations, 70% to 80% of gallstones
are cholesterol, and black pigment stones account for most of the remain-
ing 20% to 30%. An important biliary precipitate in gallstone pathogenesis is
biliary “sludge,” which refers to a mixture of cholesterol crystals, calcium bili-
rubinate granules, and mucin gel matrix. Biliary sludge has been observed
clinically in prolonged fasting states and with the use of long-term total par-
enteral nutrition (TPN). Both of these conditions also are associated with
gallstone formation. The finding of macromolecular complexes of mucin
and bilirubin, similar to biliary sludge in the central core of most cholesterol
gallstones, suggests that sludge may serve as the nidus for gallstone growth.
Cholesterol Gallstones
The pathogenesis of cholesterol gallstones is multifactorial but may be con-
sidered to involve four factors: (a) cholesterol supersaturation in bile, (b)
crystal nucleation, (c) gallbladder dysmotility, and (d) gallbladder absorp-
tion/secretion. For many years, gallstones were thought to result primar-
ily from a defect in the hepatic secretion of biliary lipids. More recently,
gallbladder motor and mucosal functions also have been demonstrated to
play key roles in gallstone formation. Cholesterol solubility depends on the
relative concentration of cholesterol, bile salts, and phospholipids. Present
theory suggests that in states of excess cholesterol production, large cho-
lesterol–phospholipid vesicles also exceed their capability to transport
cholesterol, and crystal precipitation occurs. These c holesterol-rich vesicles
279
0002104580.INDD 279 7/14/2014 7:54:59 PM

aggregate to form large multilamellar liquid vesicles that then precipitate

cholesterol monohydrate crystals. Several pronucleating factors includ-
ing mucin glycoproteins, immunoglobulins, and transferrin accelerate the
precipitation of cholesterol in bile.
For gallstones to cause clinical symptoms, they must obtain a size suf-
ficient to produce a mechanical injury to the gallbladder or obstruction of
the biliary tree. Growth of stones may occur in two ways: (a) progressive
enlargement of individual crystals or stones or (b) fusion of individual crys-
tals or stones to form a larger conglomerate (Fig. 23.1A). In addition, defects
FIGURE 23.1 A. Pure cholesterol gallstones. B. Cholesterol gallstones

with calcium.
(Continued)
002104580.INDD 280 7/14/2014 7:55:00 PM

Chapter 23 / Gallbladder Disease 281
FIGURE 23.1 (Continued) C. Black pigment stones. D. Brown pigment stones.
in gallbladder motility increase the residence time of bile in the gallbladder,

thereby playing a role in stone formation. Gallstone formation occurs in
clinical states with gallbladder stasis, as seen with prolonged fasting, the
use of long-term parenteral nutrition, after vagotomy, in diabetic patients,
and in patients with somatostatin-producing tumors or in those receiving
long-term somatostatin therapy.
The gallbladder is a very effective absorptive organ, and a normal func-
tion of the gallbladder is to concentrate and acidify bile. For many years,
three factors have been thought to be key in cholesterol gallstone patho-
genesis: cholesterol supersaturation, cholesterol crystallization, and biliary
0002104580.INDD 281 7/14/2014 7:55:03 PM

motility. However, a fourth factor, gallbladder absorption/secretion, also is

key to gallstone formation. Alterations in sodium, chloride, bicarbonate,
and water absorption/secretion alter the milieu for both cholesterol crystal
formation and for calcium precipitation with various anions (Fig. 23.1B).
Pigment Gallstones
Precipitation of calcium with bilirubin, carbonate, phosphate, or palmitate
as insoluble calcium salts serves as a nidus for cholesterol stone formation.
Furthermore, calcium bilirubinate and calcium palmitate also form major
components of pigment gallstones. Pigment gallstones are classified as
either black or brown pigment stones. Black pigment stones are typically
tarry and frequently are associated with hemolytic conditions or cirrhosis
(Fig. 23.1C). In hemolytic states, the bilirubin load and concentration of
unconjugated bilirubin increases. These stones usually are not associated
with infected bile and are located almost exclusively in the gallbladder.
In contrast to black pigment stones, brown pigment stones are earthy
in texture and are typically found in the bile ducts, especially in Asian
populations (Fig. 23.1D). Brown stones often contain more cholesterol and
calcium palmitate than black stones and occur as primary common duct
stones in Western patients with disorders of biliary motility and associated
bacterial infection. In these settings, bacteria producing slime and bacte-
ria containing the enzyme glucuronidase cause enzymatic hydrolysis of
soluble conjugated bilirubin glucuronide to form free bilirubin, which then
precipitates with calcium.
Prevalence
Gallstones are uncommon in patients younger than age 20 years, but
a sharp increase is noted, especially in women, with each decade to
approximately 70 years. Approximately 20% of women and 10% of Western
men have stones by age 60 years. In certain populations, such as Native
Americans, the incidence is extremely high, especially in women. In
Chileans and Bolivians of Indian ancestry, gallstones also are very com-
mon, and they are associated with a high incidence of gallbladder can-
cer. In the United States, the prevalence of stones is highest in Mexican
American women (26%), and the prevalence in white women (17%) is
higher than in black women (14%).
Risk Factors
Gallstones are more common in women, especially those who are obese,
have had multiple pregnancies, are taking birth control pills, are undergo-
ing rapid weight loss, or have elevated serum triglyceride levels. Diet plays
an important role in cholesterol supersaturation, and these gallstones do
not form in vegetarians. Cholesterol gallstones are common in populations
consuming a Western diet, which is relatively high in overall calories as well
as animal fats and carbohydrates. Diabetic patients also have an increased
incidence of gallstones, which may be caused, in part, by alterations in gall-
bladder motor function and/or absorption/secretion. Gallstones also are
known to occur more frequently in certain families. Current theory suggests
that approximately 30% of the risk for gallstone formation is hereditary,
whereas 70% is environmental, with diet being the primary environmen-
tal factor. As mentioned previously, prolonged fasting, TPN, ileal resection,
vagotomy, hemolytic states, and cirrhosis are additional risk factors, and
many of these factors lead to black pigment stone formation. Finally, bile
duct stasis, as occurs with biliary strictures, congenital cysts, chronic pan-
creatitis, sclerosing cholangitis, and perivaterian duodenal diverticula, is
the primary risk factor for brown pigment stone formation.
002104580.INDD 282 7/14/2014 7:55:03 PM

TABLE
23.1 Natural History of Gallstones
Symptoms Requiring Complications

Degree of Symptoms Cholecystectomy of Gallstones
Asymptomatic 1%–2%/y 1%–2%/y
Mild symptoms 6%–8%/y 1%–3%/y
Symptomatic 5%–30%/y 7%/y
Natural History
An understanding of the natural history of gallstone disease is necessary
for the appropriate management of patients with cholelithiasis. The pres-
ence or absence of symptoms remains the most important factor in the
determination of the natural history of gallstones. Gallstone disease can be
considered as a spectrum of clinical entities that includes asymptomatic
gallstones, symptomatic gallstones, and complicated gallstone disease. The
complications of gallstone disease include (a) acute cholecystitis, (b) cho-
ledocholithiasis with or without cholangitis, (c) gallstone pancreatitis, (d)
gallstone ileus, and (e) gallbladder carcinoma.
Asymptomatic gallstones often are discovered at the time of laparot-
omy or during abdominal imaging for nonbiliary disease. The vast majority
of patients with gallstones are asymptomatic. These stones remain in the
gallbladder and do not obstruct the cystic duct. As a result, the gallbladder
fills and empties normally, and the gallstones remain silent. However, asymp-
tomatic gallstones can progress to symptomatic disease. Symptomatic gall-
stones usually present with what is termed biliary colic, right upper quadrant
or epigastric abdominal pain that typically develops postprandially and may
be associated with nausea and vomiting. The pain results from the impaction
of a gallstone at the neck of the gallbladder or cystic duct.
Studies that have followed asymptomatic patients have shown that 20%
to 30% of patients become symptomatic within 20 years. Approximately 1%
to 2% of asymptomatic individuals with gallstones per year develop serious
symptoms or complications related to their gallstones (Table 23.1). The lon-
ger stones remain silent, the less likely symptoms are to develop. In addition,
almost all patients will develop symptomatic disease before developing one
of the complications of gallstones. Therefore, prophylactic cholecystectomy
generally is not indicated in patients with asymptomatic gallstones.
In select groups of patients, however, prophylactic cholecystectomy
should be considered (Table 23.2). Children with gallstones almost always
TABLE
23.2 Indications for Prophylactic Cholecystectomy
Pediatric gallstones
Congenital hemolytic anemia
Gallstones >2.5 cm in diameter
Calcified (porcelain) gallbladder
Long common channel
Bariatric surgery
Incidental gallstones found during intra-abdominal surgery
Solid organ transplantation
No access to medical care
0002104580.INDD 283 7/14/2014 7:55:04 PM

develop symptoms and should be considered for early cholecystectomy. In

patients with sickle cell disease, cholecystitis can precipitate a crisis with sub-
stantial operative risks. Therefore, these patients are best treated with elective
cholecystectomy. A nonfunctioning gallbladder usually indicates advanced
disease with more than 25% of these patients developing symptoms that
require cholecystectomy. Large gallstones (>2.5 cm) are more frequently asso-
ciated with acute cholecystitis and gallbladder carcinoma, and prophylactic
cholecystectomy also may be indicated in these patients. The presence of a
porcelain gallbladder (calcified gallbladder wall) is associated with a 5% to 10%
risk of malignant transformation, and prophylactic cholecystectomy has been
recommended for these patients. A long common channel between the bile
and pancreatic ducts also is a significant risk for gallbladder cancer, and these
patients should undergo prophylactic cholecystectomy. In obese patients in
whom gallstones have already developed and who undergo bariatric surgery,
cholecystectomy may be indicated because symptoms are likely to develop
and may be difficult to distinguish from those caused by complications of the
operation. Prophylactic cholecystectomy adds minimal morbidity and mortal-
ity risks to most bariatric operations and is clearly indicated in patients with
gallstones. Finally, acute cholecystitis is a potentially life-threatening condition
in immunosuppressed patients. For this reason, prophylactic cholecystectomy
has been recommended prior to solid organ transplantation.
Patients with mild symptoms (intermittent biliary colic) are at higher
risk for developing gallstone-related complications than asymptomatic
patients who have gallstones (Table 23.1). Approximately 1% to 3% of mildly
symptomatic patients per year will develop gallstone-related complica-
tions, and at least 6% to 8% per year will require a cholecystectomy to man-
age their gallbladder symptoms. The diagnosis of symptomatic gallstones
requires the presence of characteristic symptoms and the documentation
of gallstones on diagnostic imaging.
Cholecystitis
Chronic Cholecystitis
The term chronic cholecystitis implies an ongoing or recurrent inflam-
matory process involving the gallbladder. In more than 90% of patients,
gallstones are the causative factor and lead to recurrent episodes of
cystic duct obstruction manifest as biliary pain or colic. Over time, these
recurrent attacks can result in scarring and a nonfunctioning gallbladder.
Histopathologically, chronic cholecystitis is characterized by an increase in
subepithelial and subserosal fibrosis and a mononuclear cell infiltrate. The
primary symptom associated with chronic cholecystitis or symptomatic
cholelithiasis is pain, often labeled biliary colic. The pain is usually located
in the right upper quadrant and/or epigastrium and frequently radiates to
the right upper back, right scapula, or between the scapulae. Other symp-
toms such as nausea and vomiting often accompany each episode. The
physical examination is usually completely normal in patients with chronic
cholecystitis, particularly if they are pain-free. During an episode of biliary
colic, mild right upper quadrant tenderness may be present. Laboratory
values such as serum bilirubin, transaminases, and alkaline phosphatase
usually are normal in patients with uncomplicated gallstones.
Acute Cholecystitis
Acute cholecystitis is the most common complication of gallstones occurring
in 15% to 20% of patients with symptomatic disease. As in biliary colic, acute
cholecystitis results from a stone impaction at the gallbladder–cystic duct
junction (Fig. 23.2A). The extent of inflammation and the progression of acute
002104580.INDD 284 7/14/2014 7:55:04 PM

FIGURE 23.2 A. Acutely inflamed, edematous gallbladder with solitary pigment stone
obstructing the cystic duct. B. Gangrenous cholecystitis with solitary black pigment stone.
cholecystitis are related to the duration and degree of obstruction. In the most
severe cases (5% to 18%), this process can lead to ischemia and necrosis of the
gallbladder wall (Fig. 23.2B). More frequently, the gallstone is dislodged, and the
inflammation gradually subsides. Acute cholecystitis is primarily an inflamma-
tion and not an infectious process with bacterial infection appearing as a sec-
ondary event. Approximately 50% of patients with acute cholecystitis will have
positive bile cultures, with Escherichia coli being the most common organism.
Patients with acute cholecystitis typically present with right upper
quadrant pain that is similar to that of biliary colic. In acute cholecysti-
tis, however, the pain is usually unremitting, may last several days, and is
0002104580.INDD 285 7/14/2014 7:55:06 PM

often associated with nausea, emesis, anorexia, and fever. On physical

examination, patients with acute cholecystitis usually have a low-grade
fever and exhibit localized right upper quadrant tenderness and guarding.
The presence of Murphy sign, an inspiratory arrest during deep palpation
of the right upper quadrant, is the classic physical finding of acute chole-
cystitis. A palpable right upper quadrant mass is appreciated in one-third
of patients and usually represents omentum that has migrated to the area
around the gallbladder in response to the inflammation. Severe jaundice is
rare, but mild jaundice may be present in up to 30% of patients. Severe jaun-
dice suggests the presence of CBD stones, cholangitis, or Mirizzi syndrome,
obstruction of the common hepatic duct by severe pericholecystic inflam-
mation resulting from impaction of a large stone in the Hartmann pouch.
Laboratory evaluation can show a mild leukocytosis (white blood cell
[WBC] count 12,000 to 15,000 cells/mm3). However, many patients have a
normal WBC. A white cell count greater than 20,000 should suggest further
complications of cholecystitis, such as gangrene, perforation, or cholangitis.
Biliary Dyskinesia
A subgroup of patients presenting with typical symptoms of biliary colic
(postprandial right upper quadrant pain, fatty food intolerance, and nau-
sea) will not have any evidence of gallstones on ultrasound examination.
The vast majority of these patients are women who frequently are over-
weight or obese. Experimental work suggests that excess fat and cytokines
in the gallbladder wall (steatocholecystitis) may be the underlying cause
for this phenomenon. Many of these gallbladders are enlarged at rest sug-
gesting that a defect in absorption/secretion may be a contributing factor.
Further investigations usually are performed in these patients to
exclude any other pathology. This workup often includes an abdominal
computed tomography (CT) scan, an esophagogastroduodenoscopy, or
even an endoscopic retrograde cholangiogram. In these patients, the diag-
nosis of biliary dyskinesia or chronic acalculous cholecystitis should be
considered. The cholecystokinin-Tc-HIDA scan has been useful in identify-
ing patients with this disorder. Cholecystokinin (CCK) is infused intrave-
nously, and the gallbladder ejection fraction (EF) is calculated. An EF of less
than 35% at 20 minutes is considered abnormal.
DIAGNOSTIC IMAGING
Abdominal X-Rays
In general, plain abdominal x-rays have a low yield in diagnosing biliary tract
problems. Plain films are most useful in diagnosing other causes of acute
abdominal pain such as a perforated viscus or a bowel obstruction. Only
approximately 15% of gallstones contain sufficient calcium to appear radi-
opaque on a plain x-ray. Rarely, abdominal films may show a calcified gall-
bladder wall or pneumobilia that may aid in the diagnosis of biliary disease.
Ultrasound
Transabdominal ultrasound is the radiologic procedure of choice for identify-
ing gallstones and bile duct dilation. Ultrasound is noninvasive, inexpensive,
and widely available. Patients should receive nothing by mouth for several
hours prior to performing an ultrasound examination so that the gallblad-
der is fully distended. Gallstones create echoes that are reflected back to the
ultrasound probe. The ultrasound waves cannot penetrate the stones, and
therefore, acoustic shadowing is seen posterior to the stones (Fig. 23.3). In
addition, gallstones that are free floating in the gallbladder will move to a
dependent position when the patient is repositioned during scanning. When
002104580.INDD 286 7/14/2014 7:55:06 PM

FIGURE 23.3 Gallbladder ultrasound with solitary stone with acoustic shadowing.
these two features are present, the accuracy of ultrasound at diagnosing

gallstones approaches 100%. Echoes without shadows may be caused by gall-
bladder polyps. However, the majority of “polyps” are really small gallstones
that are attached to the mucus gel.
Several features lower the diagnostic accuracy of ultrasound in detect-
ing gallstones. Small gallstones may not demonstrate an acoustic shadow.
A lack of fluid (bile) around the gallstones (stone impacted in the cystic
duct, gallbladder filled with gallstones) also impairs their detection. In
addition, an ileus with increased abdominal gas, as occurs with acute cho-
lecystitis, will hamper gallbladder visualization. Overall, the false-negative
rate for ultrasound in detecting gallstones is approximately 5% but may
increase to 15% with acute cholecystitis.
Cholescintigraphy
Cholescintigraphy provides a noninvasive evaluation of the liver, gallblad-
der, bile duct, and duodenum with both anatomic and functional infor-
mation. 99mTechnetium-labeled iminodiacetic acid derivatives (hepatic
2, 6-dimethyl-iminodiacetic acid [HIDA], diisopropylacetanilidoiminodi-
acetic acid, or p-isopropylacetanilido imidodiacetic acid) are injected intra-
venously, rapidly extracted from the blood, and excreted into the bile. These
radionuclide scans provide functional information about the liver’s ability to
excrete radiolabeled substances into a nonobstructed biliary tree. Uptake by
the liver, gallbladder, common bile duct (CBD), and duodenum should all be
present after 1 hour. Slow uptake of the tracer by the liver suggests hepatic
parenchymal disease. Filling of the gallbladder and CBD with delayed or
absent filling of the intestine suggests an obstruction at the ampulla.
The primary use of cholescintigraphy is in the diagnosis of acute cho-
lecystitis. Although used less frequently for this indication because of the
0002104580.INDD 287 7/14/2014 7:55:06 PM

FIGURE 23.4 HIDA scan demonstrating good hepatic uptake, a patent CBD
(CD), and visualization of the duodenum (D). This scan is “positive” for acute
cholecystitis because the gallbladder is not visualized.
availability and accuracy of ultrasound, cholescintigraphy demonstrates the

presence of cystic duct obstruction, which is invariably present in acute cho-
lecystitis. Nonvisualization of the gallbladder 1 hour after the injection of the
radioisotope with filling of the CBD and duodenum is consistent with total or
partial cystic duct obstruction and acute cholecystitis (Fig. 23.4). The sensitiv-
ity and specificity of cholescintigraphy for diagnosing acute cholecystitis are
each about 95%. False-positive results are increased in the setting of gallblad-
der stasis as in critically ill patients or in those receiving parenteral nutrition.
Computerized Tomography/Magnetic Resonance Imaging

Abdominal CT is less sensitive in diagnosing gallstones than ultrasound.
Calcified gallstones are visualized in approximately 50% of patients. The role
of CT scanning is primarily limited to the diagnosis of complications of gall-
stone disease such as acute cholecystitis (gallbladder wall thickening, peri-
cholecystic fluid), gallbladder perforations, choledocholithiasis (intrahepatic
and extrahepatic bile duct dilation), pancreatitis (pancreatic edema and
inflammation), and gallbladder cancer (Fig. 23.5). Magnetic resonance imag-
ing (MRI) also has been shown to be highly sensitive in the diagnosis of both
gallstones and common duct stones when T2-weighted images are obtained.
Both MRI and CT are much more expensive than ultrasound and, therefore,
are not cost-effective for the initial evaluation of gallbladder disease.
MANAGEMENT
The operative management of gallstones has been the standard of
care over the past 140 years. Cholecystectomy is the most common
002104580.INDD 288 7/14/2014 7:55:07 PM

FIGURE 23.5 A. CT scan demonstrating a thickened gallbladder wall

secondary to cancer. Note the low-density area in the adjacent liver due to
metastatic disease. B. Laparoscopic appearance of the gallbladder in the
same patient. Note that the gallbladder wall is white with neovascularization.
gastrointestinal operation performed in the United States. Since the

introduction of laparoscopic cholecystectomy in the late 1980s, the num-
ber of cholecystectomies performed in the United States has increased
from approximately 500,000 to 750,000 per year. Symptomatic choleli-
thiasis is the main indication for laparoscopic cholecystectomy. Relative
contraindications to laparoscopic cholecystectomy include the inability
0002104580.INDD 289 7/14/2014 7:55:08 PM

of the patient to withstand a general anesthetic, severe bleeding disor-

ders, and end-stage liver disease. In addition, patients with severe chronic
obstructive pulmonary disease or congestive heart failure may not tol-
erate the pneumoperitoneum required for performing laparoscopic sur-
gery. Morbid obesity, once thought to be a relative contraindication to the
laparoscopic approach, is not associated with a higher conversion rate.
Longer trocars and instruments and an increase in intra-abdominal pres-
sure may be helpful in these patients. Prior upper abdominal surgery may
increase the difficulty but rarely precludes laparoscopic cholecystectomy.
Elective laparoscopic cholecystectomy also has been completed safely in
patients with well-compensated cirrhosis (Child class A and B), although
difficulty retracting the firm liver and increased bleeding from collaterals
have been noted.
Conversion to an open cholecystectomy is required in less than 3%
of patients undergoing laparoscopic cholecystectomy for chronic cholecys-
titis. Conversion rates are increased in male and elderly patients. Elective
laparoscopic cholecystectomy can be safely performed as an outpatient
procedure (Fig. 23.6). Among patients selected for outpatient management,
75% to 95% of patients can be successfully discharged the same day. Factors
contributing to overnight admission include uncontrolled pain, nausea and
vomiting, urinary retention, and cases completed late in the day.
Acute Cholecystitis
Once the diagnosis of acute cholecystitis is made, the patient should be
given nothing by mouth, and intravenous hydration should begin. A naso-
gastric tube is placed if persistent nausea and vomiting or abdominal
FIGURE 23.6 Laparoscopic appearance of a minimally invasive diseased gallbladder, which

contained symptomatic gallstones.
002104580.INDD 290 7/14/2014 7:55:09 PM

istention are present. In almost all cases, broad-spectrum antibiotics,

d
such as a broad-spectrum penicillin, should be started and maintained into
the immediate postoperative period. Parenteral analgesia also should be
administered. Unfortunately, narcotics increase biliary pressure, whereas
nonsteroidal analgesics, which inhibit prostaglandin synthesis, reduce gall-
bladder mucin production and, therefore, reduce pressure and pain.
The treatment of choice for acute cholecystitis is cholecystectomy.
Open cholecystectomy had been the standard treatment for acute chole-
cystitis for many years. Initially, acute cholecystitis was felt to be a contra-
indication to laparoscopic cholecystectomy. As experience has increased,
however, laparoscopic cholecystectomy clearly can be performed safely in
the setting of acute cholecystitis. However, the conversion rate in the set-
ting of acute cholecystitis (10% to 20%) is higher than with chronic chole-
cystitis. In prospective, randomized trials, length of hospital stay and time
to return to work have been lower in patients undergoing early as opposed
to late laparoscopic cholecystectomy. In addition, randomized trials, popu-
lation studies, and cost–utility analyses have shown that early laparoscopic
cholecystectomy (within 3 days of symptom onset) can be accomplished
with a similar morbidity and mortality rate and lower cost than with
delayed cholecystectomy. No significant differences were observed in the
conversion rate to open cholecystectomy among patients undergoing early
cholecystectomy versus those managed with delayed surgery. In addition,
approximately 20% of patients in whom delayed surgery is planned require
an operation during the initial admission or before the end of the planned
cooling off period. Moreover, a recent Canadian propensity score analysis
demonstrated that patients undergoing early surgery had a lower incidence
of major bile duct injuries as well as reduced mortality.
In certain high-risk patients whose medical condition precludes cho-
lecystectomy, a cholecystostomy can be performed for acute cholecystitis.
In this setting, percutaneous gallbladder drainage usually can be accom-
plished. In most cases, prompt improvement is seen after gallbladder
drainage and appropriate antibiotics. However, patients must be observed
closely, and if improvement does not occur within 24 hours, laparotomy is
indicated. Failure to improve after percutaneous cholecystostomy is usu-
ally caused by gangrene of the gallbladder or perforation. After the acute
episode resolves, the patient can undergo either cholecystectomy or per-
cutaneous stone extraction and removal of the cholecystostomy tube. The
latter is an option in elderly or debilitated patients for whom a general anes-
thetic is contraindicated.
Severe Cholecystitis
Several complications of acute cholecystitis are recognized in clinical prac-
tice. These complications include empyema of the gallbladder, emphysema-
tous cholecystitis, gangrene, perforation, and cholecystoenteric fistula. All
of these complications are associated with significant morbidity and mor-
tality and, therefore, require prompt surgical intervention.
Empyema
Gallbladder empyema is a rare advanced stage of cholecystitis with bacterial
invasion of the gallbladder and actual pus in the lumen (Fig. 23.7). Patients
present with severe right upper quadrant pain, high-grade fever, rigors, and
significant leukocytosis and mild elevations of bilirubin (2 to 3 mg/dL).
Sepsis, including cardiovascular collapse, may be seen. Treatment consists
of broad-spectrum antibiotics, including anaerobic coverage, and emergent
cholecystectomy or cholecystostomy. However, cholecystostomy may not
be adequate to relieve sepsis in patients with empyema of the gallbladder.
0002104580.INDD 291 7/14/2014 7:55:09 PM

FIGURE 23.7 Empyema of the gallbladder. Note the thickened, inflamed wall and pus
around the multiple pigment stones.
Emphysematous Cholecystitis
Emphysematous cholecystitis is another uncommon (1%) form of acute
cholecystitis, which develops more frequently in males and in patients
with diabetes mellitus. Only half of these patients have gallstones, but the
majority (75%) have a gangrenous gallbladder with (20%) or without perfo-
ration (Fig. 23.8A). Severe right upper quadrant pain and generalized sepsis
are frequently present. Abdominal films or CT scans may demonstrate air
within the gallbladder wall or lumen (Fig. 23.8B). Prompt antibiotic therapy
to cover the common biliary pathogens, including E. coli, Enterococcus,
Klebsiella, as well as Clostridia species, and emergency cholecystectomy are
appropriate treatments.
Gangrene/Perforation
Gangrene of the gallbladder, another rare complication, occurs when
the wall becomes ischemic and leads to perforation. Again, only 50% of
these patients have gallstones, and many are diabetic. Gangrene is more
likely in a number of clinical situations including ruptured abdominal
aortic aneurysms, following cardiac surgery, and in patients with burns,
trauma, and a long intensive care unit stay as well as in patients requiring
TPN. Gallbladder perforation occurs in 5% to 10% of patients with acute
cholecystitis and can be categorized as either localized or free. Localized
perforation generally results in the formation of a pericholecystic abscess
as the omentum walls off the perforation and limits it to the right upper
quadrant (Fig. 23.9A). Free perforation is less frequent (1% of cases) and
occurs if the omentum is unable to wall off the inflammatory process. Free
perforation results in the spilling of bile into the peritoneal cavity and a
generalized peritonitis. Perforation should be suspected if the patient’s
clinical course deteriorates. Evidence for perforation includes an increase
in pain and tenderness, fever and chills, elevation in WBC count, and
hypotension. These patients require aggressive fluid resuscitation, anti-
biotics, and emergent operative exploration (Fig. 23.9B). Cholecystostomy
usually will not be adequate therapy for patients with gangrene or perfo-
ration of the gallbladder.
002104580.INDD 292 7/14/2014 7:55:10 PM

FIGURE 23.8 A. Plain abdominal x-ray with a round gas pattern

(arrows) in the right upper quadrant. B. Open cholecystectomy in
this patient demonstrating a gangrenous gallbladder fundus.
Cholecystoenteric Fistula
In 1% to 2% of patients with acute cholecystitis, the gallbladder will
perforate into an adjacent hollow viscus. The duodenum (75% to 80%)
and the hepatic flexure of the colon (15% to 20%) are the most common
sites. Generally, after the fistula forms, the episode of acute cholecystitis
resolves as the gallbladder spontaneously decompresses. If a large gall-
stone passes from the gallbladder into the small intestine, a mechani-
cal bowel obstruction may result, which is termed gallstone ileus.
Gallstone ileus occurs in 10% to 15% of patients with a cholecystoenteric
0002104580.INDD 293 7/14/2014 7:55:11 PM

FIGURE 23.9 A. CT scan demonstrating a very enlarged gallbladder with

a thickened wall. B. At surgery, this patient had a localized pericholecystic
abscess. This picture demonstrates the area after gallbladder removal
and abscess evacuation.
fistula. Patients with gallstone ileus present with signs and symptoms
of intestinal obstruction—nausea, vomiting, and abdominal pain. The
pain may be episodic and recurrent as the impacted stone temporarily
impacts in the gut lumen and then dislodges and moves distally (tum-
bling obstruction). A history of gallstone-related symptoms (right upper
quadrant pain) may only be present in 50% of these patients. Abdominal
films will demonstrate small bowel distension and air–fluid levels and
may give additional clues to the source of the obstruction (pneumobi-
lia or a calcified gallstone d
istant from the gallbladder) (Fig. 23.10A).
002104580.INDD 294 7/14/2014 7:55:12 PM

FIGURE 23.10 A. Gallstone ileus. Note air in the biliary tree (upper right), air–fluid
levels (middle), and large calcified gallstone (lower right). B. At surgery, in this
patient, a large gallstone was found to be obstructing the distal ileum.
The site of obstruction is most frequently in the narrowest part of the

small intestine (ileum) or large intestine (sigmoid colon).
The initial management of gallstone ileus includes relieving the
obstruction. Most frequently, this goal can be achieved by removing the gall-
stone through an enterotomy (Fig. 23.10B). Additional gallstones should be
0002104580.INDD 295 7/14/2014 7:55:13 PM

sought as recurrent obstruction has been reported in up to 10% of patients

with gallstone ileus. Takedown of the biliary–enteric fistula and cholecys-
tectomy may be warranted as recurrent cholecystitis and cholangitis are
common in patients with a biliary–enteric fistula, and gallbladder cancer
has been reported in 15% of these patients. However, in patients with a
significant inflammatory process in the right upper quadrant or who are
unstable to withstand a prolonged operative procedure, the fistula should
be addressed at a second laparotomy.
Mirizzi Syndrome
The Mirizzi syndrome is a rare form of gallbladder disease that usually
presents with pain, jaundice, and/or cholangitis. Two types have been
described. In type I, a large stone becomes impacted in the cystic duct on
the Hartmann pouch (Fig. 23.11A). In these patients, partial cholecystec-
tomy with “repair” of the bile duct with or without a T-tube may be pos-
sible. In type II patients, the stone erodes into the common hepatic and
CBD (Fig. 23.11B). Usually, a Roux-en-Y hepaticojejunostomy is required to
repair this situation.
OUTCOMES
Serious complications of laparoscopic cholecystectomy are rare, and the
mortality rate is less than 0.3%. As cholecystectomy rates have risen, how-
ever, the total number of deaths has not decreased. Although the increase
in cholecystectomy rates means that the total number of deaths from
cholecystectomy has not decreased, an individual patient’s risk for death
is smaller. The single greatest problem in laparoscopic cholecystectomy
is biliary injury. The incidence of major bile duct injury following laparo-
scopic cholecystectomy is between 0.3% and 0.6%, but if all bile leaks are
considered, the injury rate in these reports ranges from 0.6% to 1.5%, which
is three to four times the injury rate at open surgery. Major vascular inju-
ries to the hepatic arteries, especially the right hepatic artery, may occur
in association with biliary injuries and sometimes lead to intraoperative
blood loss. Vasculobiliary injuries may complicate the management and
prognosis of these patients.
Spillage of stones into the peritoneal cavity during laparoscopic
cholecystectomy occurs in 10% or more of cases. Leaving stones in the
peritoneal cavity may not be innocuous. Intra-abdominal abscess, sub-
cutaneous abscess, and later discharge of stones through the abdomi-
nal wall or through the lung and trachea have all been described. Every
attempt should be made to remove spilled stones by picking and irri-
gating them out. Clearance is usually quite successful with the use of
retractors to lift the liver and the 30-degree laparoscope, which allows
the depths of the recess between the liver and kidney to be visualized.
Laparoscopic ultrasonography may be useful to detect these stones.
Large stones or massive spills should be cleaned up by laparotomy if nec-
essary. If concern exists that stones have been left behind, the patient
should be informed.
A gallbladder containing an unsuspected cancer is excised one to
three times per 1,000 laparoscopic cholecystectomies. Older patients
and those requiring an open cholecystectomy (planned or converted) are
at greater risk. Therefore, a good practice is to open the gallbladder and
inspect it and to obtain frozen sections if a suspect lesion is observed. If
cancer is suspected, the gallbladder should be extracted in an impermeable
bag. If a cancer is discovered, further surgery may be indicated.
002104580.INDD 296 7/14/2014 7:55:13 PM

FIGURE 23.11 A. Mirizzi type II schematic. B. Mirizzi type II cholan-

giogram with a large stone eroded into the CBD.
0002104580.INDD 297 7/14/2014 7:55:13 PM

The long-term results of laparoscopic cholecystectomy in a ppropriately

selected patients with chronic cholecystitis are excellent. More than 90% of
patients with typical biliary pain and gallstones are symptom-free follow-
ing cholecystectomy. Results of open cholecystectomy are slightly worse
because a small percentage of patients have persistent incisional pain. For
patients with atypical symptoms or painless dyspepsia ( fatty food intoler-
ance, flatulence, belching, or bloating), the percentage of patients experi-
encing relief of symptoms decreases. Between 80% and 90% of patients with
a low gallbladder ejection fraction and symptoms of biliary colic will be
asymptomatic or improved by cholecystectomy. Most of these patients have
histopathologic evidence of chronic cholecystitis. Over the past 15 years,
the percentage of patients undergoing cholecystectomy for acalculous cho-
lecystitis (biliary dyskinesia) in the United States has increased from less
than 5% to more than 20% of patients having their gallbladder removed.
This change parallels the obesity epidemic and may be due to abnormal
gallbladder emptying as well as absorption/secretion due to fat within the
gallbladder wall (cholecystosteatosis).
Acute Cholecystitis
In general, the results of patients undergoing cholecystectomy for acute
cholecystitis also are quite good. As outlined above, early surgery is recom-
mended for most patients with acute cholecystitis and is associated with
fewer bile duct injuries and less mortality. On the other hand, in the small
subset of patients with severe cholecystitis, the risk of hospital mortality
is dramatically increased. Patients with emphysematous cholecystitis and
those with gangrene or perforation of the gallbladder have a 10% to 15%
mortality. Severe sepsis with associated organ failure is clearly life threaten-
ing in older, frail patients as well as those with ruptured aneurysms, recent
cardiac surgery, burns, or trauma.
CONCLUSIONS
Gallbladder disease is a very common health care problem. Considerable
knowledge exists regarding the etiology and pathogenesis of gallstones and
biliary dyskinesia. However, strategies to prevent stones, other than diet,
have not been developed. In addition, nonoperative methods to dissolve or
remove gallstones have not been successful. The fact that laparoscopic cho-
lecystectomy is quite safe and that the majority of patients undergoing this
procedure can be managed as an outpatient has thwarted nonoperative
challenges. In addition, with laparoscopic cholecystectomy, patients return
to work quickly, and in the vast majority, the long-term quality of life is
quite good. Moreover, following the introduction of laparoscopic cholecys-
tectomy, the percentage of patients presenting with acute cholecystitis and
choledocholithiasis has reduced because the threshold for patients to have
their gallbladder removed has lowered. However, approximately 8% to 10%
of patients still present with complex gallbladder disease that may require
an open cholecystectomy and/or more complex biliary procedures.
Suggested Readings
Al-Azzawi HH, Nakeeb A, Saxena R, et al. Cholecystosteatosis: an explanation for
increased cholecystectomy rates. J Gastrointest Surg 2007;11(7):835–843.
Doty JE, Pitt HA, Kuchenbecker SL, et al. Impaired gallbladder emptying prior to gall-
stone formation in the prairie dog. Gastroenterology 1983;85,168–174.
Friedman GD, Raviola CA, Fireman B. Prognosis of gallstones with mild or no symptoms:
25 years of follow-up in a health maintenance organization. J Clin Epidemiol 1989;42:
127–134.
002104580.INDD 298 7/14/2014 7:55:13 PM

Gutt CN, Encke J, Koninger J, et al. Acute cholecystitis: early versus delayed cholecys-
tectomy: a multicenter randomized trial. Ann Surg 2013;258:385–397.
de Mestral C, Rostein O, Laupacis A, et al. Comparative operative outcomes of early
and delayed cholecystectomy for acute cholecystitis. Ann Surg 2014;259:10–15.
Nakeeb A, Comuzzie AG, Martin L, et al. Gallstones: genetics versus environment.
Ann Surg 2002;235:842–847.
Pitt HA. Patient value is superior with early surgery for acute cholecystitis. Ann Surg
2014;259;16–17.
Ravikumar R, Williams JG. The operative management of gallstone ileus. Ann R Coll
Surg Engl 2010;92:279–284.
Schmidt GH, Hausken T, Glambek I, et al. A 24-year controlled follow-up of patients
with silent gallstones showed no long-term risk of symptomatic or adverse events
leading to cholecystectomy. Scan J Gastroenterol 2001;46:949–54.
Stephen AE, McFadden DW, Cortina GR, et al. Porcelain gallbladder is not associated
with gallbladder carcinoma. Am Surgeon 2001;67:7–11.
Stinton LM, Myers RP, Shaffer EA. Epidemiology of gallstones. Gastroenterol Clin
North Am 2010;39:157–175.
Strasberg SM, Brunt LM. Rationale and use of the critical view in laparoscopic chole-
cystectomy. Am J Surg 1993;165:655–60.
Strasberg SM. Clinical practice. Acute calculus cholecystitis. N Engl J Med 2008;358:
2804–2811.
Swartz-Basile DA, Lu D, Basile DP, et al. Leptin regulates gallbladder genes related to
absorption and secretion. Am J Physiol 2007;293:84–89.
Takada T, Strasberg SM, Solomkin JS, et al. Updated Tokyo guidelines for the man-
agement of acute cholecystitis and cholangitis. J Hepatobiliary Pancreat Sci
2013;20:1–7.
0002104580.INDD 299 7/14/2014 7:55:13 PM

24 Sphincter of Oddi Dysfunction
Evan L. Fogel
INTRODUCTION
Since its original description by Ruggero Oddi in 1887, the sphincter of Oddi
(SO) has been the subject of much study and controversy. Its very existence
as a distinct entity has been disputed. It is therefore not surprising that the
clinical syndrome of sphincter of Oddi dysfunction (SOD) and its therapy
are controversial areas. Nevertheless, SOD is commonly diagnosed and
treated by physicians. This chapter reviews the epidemiology and clinical
presentation of SOD and currently available diagnostic and therapeutic
modalities.
Postcholecystectomy pain resembling the patient’s preoperative bili-
ary colic occurs in at least 10% to 20% of patients. These patients should
have appropriate noninvasive and invasive (when indicated) evaluation
to rule out common bile duct stones, tumors, or strictures near the cho-
lecystectomy site. The residual group of patients has a high frequency of
SOD. SOD refers to an abnormality of SO contractility. It is a benign, non-
calculous obstruction to flow of bile or pancreatic juice through the pan-
creaticobiliary junction (i.e., the SO) resulting from a dyskinetic or stenotic
sphincter. SOD may be manifested clinically by “pancreaticobiliary” pain,
pancreatitis, abnormal liver tests, or abnormal pancreatic enzymes. SO
dyskinesia refers to a motor abnormality of the SO, which may result in a
hypotonic sphincter but, more commonly, causes a hypertonic sphincter.
In contrast, SO stenosis refers to a structural alteration of the sphincter,
probably from an inflammatory process, with subsequent fibrosis. Because
it is often impossible to distinguish patients with SO dyskinesia from those
with SO stenosis, the term SOD has been used to incorporate both groups
of patients. A clinical classification system has been developed for patients
with suspected biliary or pancreatic SOD based on clinical history, labora-
tory results, and endoscopic retrograde cholangiopancreatography (ERCP)
findings (Tables 24.1 and 24.2). A variety of less accurate terms—such as
papillary stenosis, ampullary stenosis, biliary dyskinesia, and postchole-
cystectomy syndrome—are listed in the medical literature to describe this
entity. The latter term is somewhat of a misnomer because SOD may clearly
occur with an intact gallbladder.
EPIDEMIOLOGY
SOD most commonly occurs in middle-aged females, although patients
of any age or sex may be affected. Although SOD typically is seen in the
postcholecystectomy state, it may occur with the gallbladder in situ. The
epidemiology of SOD is unclear due to a paucity of population-based data
and the considerable variation that exists in currently published literature,
including patient selection criteria, definition of SOD used, and whether
or not one or both sphincter segments are studied by sphincter of Oddi
300
002086388.INDD 300 7/14/2014 4:59:42 PM

Chapter 24 / Sphincter of Oddi Dysfunction 301
TABLE
Modified Milwaukee Classification for Biliary Sphincter
24.1
Biliary Type 1
of Oddi Dysfunction, Postcholecystectomy
Patients with biliary-type pain, abnormal ALT or alkaline phosphatase greater

than two times normal, and dilated common bile duct (CBD) >12 mm diameter
Biliary Type II
Patients with biliary-type pain and one of either abnormal ALT or alkaline
phosphatase greater than two times normal OR dilated CBD >12 mm
Biliary Type III
Patients with only biliary-type pain and no other abnormalities
manometry (SOM). Eversman et al. performed SOM of both the biliary and
pancreatic sphincter segments in 360 patients with intact sphincters. In
this series, 19% had abnormal pancreatic basal sphincter pressure alone,
11% had abnormal biliary basal sphincter pressure, and 31% had abnormal
basal sphincter pressure in both segments (total 61% with abnormal SOM).
A more recent 14-year review of patients undergoing evaluation with SOM
at our institution identified SOD in 65% of patients. These and other stud-
ies highlight the need to evaluate both the bile duct and pancreatic duct
during SOM. Sphincter dysfunction may also cause recurrent pancreatitis,
and manometrically documented SOD has been reported in 15% to 72% of
patients previously labeled as having idiopathic pancreatitis.
PATHOGENESIS AND PATHOLOGY

The SO is a complex of smooth muscles that surrounds the terminal com-
mon bile duct, ventral pancreatic duct, and the common channel (ampulla
of Vater) if present (Fig. 24.1). Its primary role is to regulate bile and pan-
creatic juice flow and to prevent reflux of duodenal contents into the sterile
biliary and pancreatic systems. The SO has both a variable basal pressure
and phasic contractile activity, which are under both neural and hormonal
control. Patients with type I SOD are thought to have a stenotic sphincter
rather than a sphincter in spasm, as pathologic series of sphincteroplasty
resection specimens have shown significant inflammation, reactive mus-
cular hypertrophy, and fibrosis within the papillary zone in 60% of patients.
These pathophysiologic changes at the sphincter papillary orifice are likely
responsible for ductal hypertension with resultant duct dilation, elevated
liver enzymes, and biliary-type pain in patients with type I SOD. In patients
TABLE
Modified Pancreatic Classification System for Sphincter of
24.2 Oddi Dysfunction
Pancreatic Type 1
Patients with pancreatic-type pain, abnormal amylase or lipase >1.5 times
normal on any occasion, and dilated pancreatic duct (PD) >6 mm diameter in
the head or 5 mm in the body
Pancreatic Type II
Patients with pancreatic-type pain and only one of abnormal amylase or lipase
>1.5 times normal on any occasion OR dilated PD >6 mm diameter in the head
or 5 mm in the body
Pancreatic Type III
Patients with only pancreatic-type pain and no other abnormalities
0002086388.INDD 301 7/14/2014 4:59:43 PM

FIGURE 24.1 The sphincter of Oddi.
with type II and III SOD, the mechanism of dysfunction is not related to
sphincter inflammation and scarring but is thought to be related to dys-
regulation of stimulatory and/or inhibitory factors.
How does SOD cause pain? From a theoretical point of view, this may
be related to (a) impedance of flow of bile and pancreatic juice resulting in
ductal hypertension, (b) muscular ischemia of the sphincter arising from
spastic contractions, and (c) hypersensitivity of the papilla and/or duode-
num. These mechanisms may potentially act alone or in concert to explain
the genesis of pain.
The Rome III classification system has provided diagnostic criteria for SOD
(Table 24.3). Abdominal pain is the most common presenting symptom.
The pain is usually localized to the epigastric area or right upper quadrant,
may radiate to the back or shoulder, and lasts anywhere from 30 minutes
to several hours. Pain may be precipitated by food or narcotics and often
is accompanied by nausea and vomiting. The pain may begin several years
after cholecystectomy and is usually similar in character to the pain that
initially prompted gallbladder evaluation. Alternatively, patients may have
continued pain that was not relieved by cholecystectomy. Jaundice, fever,
or chills are rarely observed. Physical examination typically is negative or
reveals only mild abdominal tenderness. The pain is not relieved by trial
medications for acid peptic disease or irritable bowel syndrome. Laboratory
abnormalities consisting of transient elevations of liver tests during epi-
sodes of pain that normalize during pain-free periods may be observed.
Patients with pancreatic SOD may present with typical pancreatic pain with
or without pancreatic enzyme elevation or recurrent pancreatitis.
The association between SOD and chronic pancreatitis is poorly
understood. It is not known whether the sphincter at times becomes dys-
functional as part of the overall scarring process or whether it has a role
in the pathogenesis of chronic pancreatitis. However, a high frequency of
basal sphincter pressure abnormalities in the pancreatic sphincter has
been identified, with 20 of 23 (87%) chronic pancreatitis patients found
to have SOD in one study. Sphincterotomy has been demonstrated to
002086388.INDD 302 7/14/2014 4:59:43 PM

TABLE
Rome III Criteria for Functional Biliary, Gallbladder,
24.3 and Sphincter of Oddi Disorders
A. Diagnostic Criteria for Functional Gallbladder and Sphincter of Oddi
Disorders
Must include episodes of pain located in the epigastrium and/or right
upper quadrant and all of the following:
1. Episodes lasting 30 min or longer
2. Recurrent symptoms occurring at different intervals (not daily)
3. The pain builds up to a steady level
4. The pain is moderate to severe enough to interrupt the patient’s daily
activities or lead to hospital visit
5. The pain is not relieved by bowel movements
6. The pain is not relieved by postural change
7. The pain is not relieved by antacids
8. Exclusions of other structural disease that would explain the symptoms
Supportive Criteria
The pain may present with one or more of the following:
1. Pain is associated with nausea and vomiting
2. Pain radiates to the back and/or right infrascapular region
3. Pain awakens from sleep in the middle of the night
B. Diagnostic Criteria for Functional Biliary Sphincter of Oddi Disorder
Must include BOTH of the following:
1. Criteria for functional gallbladder or sphincter of Oddi disorder met
2. Normal amylase/lipase
Supportive criteria:
Elevated serum transaminases, alkaline phosphatase, or conjugated bilirubin
temporally related to at least two pain episodes
C. Diagnostic Criteria for Functional Pancreatic Sphincter of Oddi Disorder
Must include BOTH of the following:
1. Criteria for functional gallbladder or sphincter of Oddi disorder met
2. Elevated amylase and/or lipase
Adapted from Behar J, Corazziari E, Guelrud M, et al. Functional gallbladder and sphincter of Oddi
disorders. Gastroenterology 2006;130:1498–1509.
improve pain in a subset of patients with chronic pancreatitis in uncon-

trolled studies.
While the diagnosis of SOD is commonly made after cholecystectomy,
SOD may also exist in the presence of an intact gallbladder. However, the
symptoms due to SOD may be indistinguishable from gallbladder-type
pain, resulting in the diagnosis of SOD being made after cholecystectomy
or less frequently after gallbladder abnormalities have been excluded.
Given the potential complications of ERCP in patients with suspected SOD
(see below), empiric cholecystectomy may be considered in select patients
as initial therapy prior to ERCP even in the setting of normal gallbladder
evaluation.
DIFFERENTIAL DIAGNOSIS
The clinical presentation of SOD may be mimicked by many organic pathol-
ogies including common bile duct stones, chronic pancreatitis, ampullary
tumors, peptic ulcer disease, mesenteric ischemia, renal colic, as well as
other functional disorders including irritable bowel syndrome, referred
musculoskeletal pain, and functional dyspepsia. Because of the 10% to
20% complication rate seen in the evaluation and therapy of patients
0002086388.INDD 303 7/14/2014 4:59:43 PM

with s uspected SOD (see below), the diagnosis should be treated as one of
exclusion, with other diagnostic possibilities initially pursued with appro-
priate testing. As well, therapeutic trials with low-risk empirical medical
therapies such as proton pump inhibitors, antispasmodics, and/or pain
modulators should be made before proceeding with ERCP and SOM.
Diagnostic Methods
Initial investigations for patients with suspected SOD should include labo-
ratory tests (liver enzymes, serum amylase and/or lipase) and abdominal
imaging (ultrasound or computed tomography (CT) scan). If at all possi-
ble, the enzyme studies should be drawn during an acute attack of pain,
although liver test abnormalities lack both sensitivity and specificity. Mild
elevations (<2× upper limit of normal) are common in SOD, whereas greater
abnormalities are more suggestive of stones, tumors, and intrinsic liver dis-
ease. Imaging of the abdomen is usually normal, but occasionally dilated
bile ducts or pancreatic ducts may be found (type I or type II patients).
More detailed structural evaluation may be obtained with EUS and MRI/
MRCP in select patients. Several noninvasive tests have been designed in an
attempt to identify those individuals with SOD. The morphine–prostigmine
provocative test (Nardi test) has been shown to have an unacceptable rate
of false-positive studies (>40% in patients with irritable bowel syndrome)
and as such is no longer recommended. Quantitative hepatobiliary scin-
tigraphy (HBS), with or without morphine provocation, may predict an
abnormal SOM and response to biliary sphincterotomy. However, abnormal
results may be found in asymptomatic controls, and HBS does not address
the pancreatic sphincter, which may be the cause of the patient’s symp-
toms. Measurement of common bile duct diameter by ultrasound after
either lipid-rich meal or secretin stimulation has also been shown to have
variable sensitivity (21% to 88%) and specificity (82% to 97%). Our group
has prospectively compared secretin-stimulated MRCP (sMRCP) to SOM.
Prediction of SOD based on sMRCP results was poor, with positive and
negative predictive values of 67% and 33%, respectively. Considering the
limitations of noninvasive testing, SOM demonstrating an elevated basal
sphincter pressure greater than 40 mm Hg (either biliary or pancreatic) is
still considered the gold standard for diagnosing SOD.
Performance of SOM
SOM is the only available method to measure SO motor activity directly
and is considered by most authorities to be the most accurate evaluation
for sphincter dysfunction. Although SOM can be performed intraopera-
tively and percutaneously, it is most commonly done in the ERCP setting.
The use of manometry to detect motility disorders of the SO is similar to its
use in other parts of the gastrointestinal (GI) tract. However, performance
of SOM is more technically demanding and hazardous, with complication
rates (in particular, pancreatitis) approaching 20% in several series. Its use,
therefore, should be reserved for patients with clinically significant or dis-
abling symptoms. One needs to appreciate, however, that SOM is not likely
an independent risk factor for post-ERCP pancreatitis when the aspirating
manometry catheter is used. It is the suspicion of SOD itself rather than the
performance of SOM that places the patient at increased risk.
The initial step in performing SOM is to administer adequate seda-
tion, which will result in a comfortable, cooperative, motionless patient.
All drugs that relax (anticholinergics, nitrates, calcium channel block-
ers, glucagon) or stimulate (narcotics, cholinergic agents) the sphincter
should be avoided for at least 8 to 12 hours prior to SOM and during the
manometric session. SOM requires selective cannulation of the bile duct
002086388.INDD 304 7/14/2014 4:59:43 PM

and/or pancreatic duct. It is preferable to perform cholangiography and/

or pancreatography prior to performance of SOM, as certain findings (e.g.,
bile duct stones) may obviate the need for SOM. Once deep cannulation
is achieved and the patient acceptably sedated, the catheter is withdrawn
across the sphincter. Ideally, both the pancreatic and bile ducts should be
studied. Current data indicate that an abnormal basal sphincter pressure
may be confined to one side of the sphincter in 35% to 65% of patients with
abnormal manometry, and thus, one sphincter segment may be dysfunc-
tional and the other normal. An abnormal basal sphincter pressure is more
likely to be confined to the pancreatic duct segment in patients with pan-
creatitis and to the bile duct segment in patients with biliary-type pain and
elevated liver function tests.
Most authorities use only the basal sphincter pressure as an indicator
of pathology of the sphincter of Oddi, with 40 mm Hg used as the upper
limits of normal for mean basal sphincter pressure (mean value plus three
standard deviations). Interobserver variability for reading SOM is minimal
when the observers are experienced in reading these tracings.
It has been questioned whether the short-term pressure recording
obtained during SOM reflect the “24-hour pathophysiology” of the sphinc-
ter, as patients with SOD may have intermittent, episodic symptoms. If the
basal sphincter pressure does vary over time, performance of SOM on two
separate occasions may lead to different results and affect therapy. Three
studies have demonstrated reproducibility of biliary SOM in 34 of 36 symp-
tomatic patients overall and 10 of 10 healthy volunteers. However, repro-
ducibility of pancreatic SOM was found in only 58% (7/12) and 40% (12/30)
of persistently symptomatic patients with previously normal SOM at two
large referral centers. Other studies have also shown that SO basal pres-
sures are not constant, perhaps due to the inherent physiologic fluctuation
of SO motor activity. Newer devices capable of portable, ambulatory, pro-
longed SOM would be of interest.
THERAPY FOR SPHINCTER OF ODDI DYSFUNCTION

The therapeutic approach in patients with SOD is aimed at reducing the
resistance to the flow of bile and/or pancreatic juice across the sphincter.
Historically, emphasis has been placed on definitive intervention (i.e., surgi-
cal sphincteroplasty or endoscopic sphincterotomy). This appears appro-
priate for patients with high-grade obstruction (i.e., type I SOD). In patients
with lesser degrees of obstruction, the clinician must carefully weigh the
risks and benefits before recommending invasive therapy. Most reports
indicate that SOD patients have a complication rate from ERCP, manom-
etry, and endoscopic sphincterotomy two to five times higher than that of
patients with ductal stones.
Medical Therapy
Medical therapy for documented or suspected SOD has received only
limited study. Because the SO is a smooth muscle structure, it is reason-
able to assume that drugs that relax smooth muscle might be an effective
treatment for SOD. Vardenafil (Levitra), an inhibitor of phosphodiesterase
type 5 and a smooth muscle relaxant used most commonly for male erec-
tile dysfunction, was found to reduce basal sphincter pressure and phasic
wave amplitude. However, this drug has not been investigated in clinical
trials. Sublingual nifedipine and nitrates have been shown to reduce the
basal sphincter pressures in asymptomatic volunteers and symptomatic
patients with SOD. In a placebo-controlled crossover trial with nifedip-
ine, 21/28 patients (75%) with manometrically documented SOD had a
0002086388.INDD 305 7/14/2014 4:59:43 PM

r eduction in pain scores, emergency room visits, and use of oral analgesics
during short-term follow-up. In a similar study, 9/12 (75%) type II SOD (sus-
pected; SOM was not done) patients improved with nifedipine. Although
medical therapy may be an attractive initial approach in patients with SOD,
several drawbacks exist. First, medication side effects may be seen in up to
one-third of patients. Second, smooth muscle relaxants are unlikely to be
of any benefit in patients with the structural form of SOD (i.e., SO stenosis),
and the response is incomplete in patients with a primary motor abnormal-
ity of the SO (i.e., SO dyskinesia). Finally, long-term outcome from medical
therapy has not been reported. Nevertheless, because of the relative safety
of medical therapy and the benign (although painful) character of SOD, this
approach should be considered in all type III and less severely symptomatic
type II SOD patients before considering the more aggressive sphincter abla-
tion therapy.
Surgical Therapy
Historically, surgery was the traditional therapy of SOD. The surgical
approach, most commonly, is a transduodenal biliary sphincteroplasty
with a transampullary septoplasty (pancreatic septoplasty). Several series
have reported that 60% to 70% of patients benefited from this therapy dur-
ing a 1- to 10-year follow-up. Predictably, patients with an elevated basal
sphincter pressure, determined by intraoperative SOM, were more likely to
improve from surgical sphincter ablation than those with a normal basal
pressure. Some reports have suggested that patients with biliary-type pain
have a better outcome than patients with idiopathic pancreatitis, whereas
others suggested no difference. However, most studies found that symptom
improvement after surgical sphincter ablation alone was relatively uncom-
mon in patients with established chronic pancreatitis.
The surgical approach for SOD has largely been replaced by endo-
scopic therapy. Patient tolerance, cost of care, morbidity, mortality, and
cosmetic results are some of the factors that favor an initial endoscopic
approach. At present, surgical therapy is reserved for patients with reste-
nosis after endoscopic sphincterotomy and when endoscopic evaluation or
therapy is not available or technically feasible. Among 68 surgical sphinc-
teroplasties done at the Medical University of South Carolina over a 5-year
period, 51 had prior endoscopic sphincterotomy and 17 had endoscopically
inaccessible papillae because of prior gastric surgery. There was a trend
toward improved outcome following surgical sphincteroplasty ( p = 0.06) in
patients who had previous gastric surgery and no prior ERCP compared to
those who had endoscopic sphincterotomy prior to their surgery. In some
centers, particularly where ERCP and SOM are not performed, operative
therapy continues to be the standard treatment of pancreatic sphincter
hypertension.
Endoscopic Therapy
Endoscopic Sphincterotomy
Endoscopic sphincterotomy is the standard therapy for patients with
SOD. Most data on endoscopic sphincterotomy relate to biliary sphincter
ablation alone. Clinical improvement after therapy has been reported to
occur in 55% to 95% of patients. These variable outcomes are reflective
of the different criteria used to document SOD, the degree of obstruc-
tion (type I biliary patients appear to have a better outcome than types
II and III), the methods of data collection (retrospective vs. prospec-
tive), and the techniques used to determine benefit. In one small study of
17 type I postcholecystectomy patients undergoing ERCP and SOM,
002086388.INDD 306 7/14/2014 4:59:43 PM

TABLE
Response to Biliary Sphincterotomy Alone in Biliary SOD
24.4 Patients
Mean Follow-up
SOD Type n Studies n Patients (Total) n Improved (%) (Months)
I 5 67 57 (85) 25.2
II 10 177 122 (69) 36.8
III 7 169 62 (37) 34.7
Adapted from Sgouros SN, Pereira SP. Systematic review: sphincter of Oddi dysfunction—non-
invasive diagnostic methods and long-term outcome after endoscopic sphincterotomy. Aliment
Pharmacol Ther 2006;24:237–246.
65% had abnormal manometry. Nevertheless, during a mean follow-up

interval of 2.3 years, all patients benefited from biliary sphincterotomy.
The results of this study suggested that because type I biliary patients
invariably benefit from biliary sphincterotomy, SOM in this patient group
not only is unnecessary but also may be misleading. However, the results
of this study have never been validated at another center. Overall, in type
I biliary SOD patients evaluated in five studies, biliary sphincterotomy led
to clinical improvement in 57/67 (85%) patients with a 2-year follow-up. In
contrast, traditional teaching, based on results of several nonrandomized
controlled trials, has emphasized that performance of SOM is highly rec-
ommended in biliary type II and is mandatory in type III patients because
clinical benefit is less certain. A total of 177 type II biliary SOD patients
were evaluated in 10 small studies and 122 (69%) improved following bili-
ary sphincterotomy alone with a mean of 3-year follow-up. Of 169 type
III patients culled from seven studies, only 62 (37%) improved follow-
ing biliary sphincterotomy, with nearly 3-year follow-up (see Table 24.4).
The addition of pancreatic sphincterotomy to biliary sphincterotomy may
offer additional benefit in patients with pancreatic sphincter hyperten-
sion, regardless of the presence of biliary SOD. Eversman et al. found that
90% of patients with persistent pain or pancreatitis after biliary sphinc-
terotomy had residual abnormal pancreatic basal pressure. Five-year
follow-up data revealed that patients with untreated pancreatic sphinc-
ter hypertension were much less likely to improve after biliary sphinc-
terotomy than patients with isolated biliary sphincter hypertension. As
shown in Table 24.5, patients with pancreatic sphincter hypertension who
fail to respond to biliary sphincterotomy can be “rescued” by u ndergoing
TABLE
Symptomatic Improvement in Pancreatic SOD Patients After
24.5
Author/Year
Pancreatic Sphincterotomy
n n (Improved %) Mean Follow-up (Months)
Pereira, 2006 13 7 (54) 30.2
Okolo, 2000 15 11 (73) 16
Elton, 1998 43 31 (72) 36.4
Soffer, 1994 25 16 (64) 13.7
Guelrud, 1995 27 22 (81) 14.7
TOTAL 123 87 (71) 23.9
Adapted from Sgouros SN, Pereira SP. Systematic review: sphincter of Oddi dysfunction—
non-invasive diagnostic methods and long-term outcome after endoscopic sphincterotomy. Aliment
Pharmacol Ther 2006;24:237–246.
0002086388.INDD 307 7/14/2014 4:59:44 PM

ancreatic sphincterotomy. Recent data from our unit examined the

p
outcome of endoscopic therapy in SOD patients with initial pancreatic
sphincter hypertension (with or without biliary sphincter hyperten-
sion). Patients were followed for a mean of 43.1 months (range 11 to 77
months); reintervention was offered for sustained or recurrent symptoms
at a median of 8 months after initial therapy. Performance of an initial
dual pancreatobiliary sphincterotomy was associated with a lower rein-
tervention rate (70/285, 24.6%) than biliary sphincterotomy alone (31/95,
33%; p < 0.05). While these results appear promising, the State of the
Science conference at the National Institutes of Health in 2002 called
for further study in the controversial type III SOD patients, as virtually
all outcome data were based on small retrospective studies. As a result,
the EPISOD (Evaluating Predictors and Interventions in Sphincter of
Oddi Dysfunction) study was undertaken. This sham-controlled prospec-
tive multicenter trial evaluated 214 type III SOD patients, randomized
to sphincterotomy or sham therapy. Importantly, while approximately
half the patients did note a reduction in pain as measured by pain dis-
ability days, those patients randomized to sham therapy did at least as
well as those treated with biliary or pancreatobiliary sphincterotomy.
Furthermore, results of SOM did not predict outcome. Additionally, this
benefit persisted at 1-year follow-up, even in the sham group. Three-year
follow-up data are currently being collected, and these results are eagerly
awaited. Further study is clearly indicated in this difficult patient popula-
tion, as the EPISOD study results suggest that the pain syndrome is not
entirely sphincter mediated.
Some authorities argue that the current SOD classification sys-
tems might not be a good predictor of outcome. In a study of 121 patients
(18 type I, 53 type II, and 50 type III) treated by biliary sphincterotomy with
(49) or without (72) pancreatic sphincterotomy, Freeman et al. reported
a good to excellent response in 69%. The response was not significantly
different between biliary types I, II, and III. The authors found that sig-
nificant predictors of a poor response to therapy were normal pancreatic
manometry, delayed gastric emptying, daily opioid use, and age less than
40. Abnormal liver function tests and dilated bile duct were not significant
predictors of outcome.
Taken as a whole, these results clearly indicate that the response rate
and enthusiasm for sphincter ablation must be correlated with patient
presentation and results of manometry (in type II patients) and balanced
against the high complication rates reported for endoscopic therapy of
SOD. As noted earlier, several prospective, multicenter studies exam-
ining risk factors for post-ERCP pancreatitis have identified suspected
(i.e., not necessarily confirmed) SOD as an independent factor by multivari-
ate analysis, with two to five times increased pancreatitis rates compared to
patients with benign or malignant biliary obstruction. These studies have
also shown that the risk of pancreatitis is intrinsic to the patient group
(“patient-related factors”) and events occurring during the procedure
(“procedure-related factors”) rather than the SOM when the SOM is per-
formed with the aspirating catheter. In a multivariate analysis, SOM has not
been shown to be a risk factor for pancreatitis.
Balloon Dilation and Stenting

Balloon dilation of strictures in the GI tract has become commonplace.
In an attempt to be less invasive and possibly preserve sphincter func-
tion, adaptation of this technique to treat SOD has been described.
Unfortunately, because of the unacceptably high complication rates,
primarily pancreatitis, this technology has little role in the primary
002086388.INDD 308 7/14/2014 4:59:44 PM

anagement of SOD. Similarly, although biliary stenting might offer

m
short-term symptom benefit in patients with SOD and predict outcome
from sphincter ablation, it too has unacceptably high complication rates
and cannot be advocated in this setting. Furthermore, pancreatic stent tri-
als are strongly discouraged due to the potential for stent-induced pancre-
atic ductal injury.
Botulinum Toxin Injection

Botulinum toxin (Botox), a potent inhibitor of acetylcholine release from
nerve endings, has been successfully applied to smooth muscle disorders of
the GI tract such as achalasia. In a preliminary clinical trial, Botox injection
into the SO resulted in a 50% reduction in the basal biliary sphincter pres-
sure and improved bile flow. This reduction in pressure may be accompa-
nied by symptom improvement in some patients. Although further study is
warranted, Botox may serve as a therapeutic trial for SOD with responders
undergoing permanent sphincter ablation. In a small series, 22 postchole-
cystectomy type III patients with manometric evidence of SOD underwent
Botox injection into the intraduodenal sphincter segment. Eleven of the 12
patients who responded to botulinum toxin injection later benefited from
endoscopic sphincterotomy, whereas only 2 of 10 patients who did not ben-
efit from Botox injection later responded to sphincter ablation. However,
such an approach does require two endoscopic procedures to achieve
symptom relief. Moreover, patients must have relatively frequent episodes
of pain to assess the benefit from Botox. Further studies are needed before
recommending this technique.
Failure to Achieve Symptomatic Improvement After Biliary Sphincterotomy

There are several potential explanations as to why patients may fail to
achieve symptom relief after biliary sphincterotomy is performed for
well-documented SOD. First, the biliary sphincterotomy may have been
inadequate, or restenosis may have occurred. Second, the importance of
pancreatic sphincter ablation is being increasingly recognized, as noted
previously. Third, patients may fail to respond to sphincterotomy because
they have chronic pancreatitis. It has been reported that SOD patients are
four times more likely to have evidence of chronic pancreatitis than those
without SOD. Although SOD appears to be associated with chronic pancre-
atitis, a causal relationship has not been proven. These patients may or may
not have abnormal pancreatograms. Intraductal pancreatic juice aspiration
after secretin stimulation (with bicarbonate determination) may help make
this diagnosis. EUS may show parenchymal and ductular changes of the
pancreas in some of these patients suggesting chronic pancreatitis. Fourth,
some patients may be having pain from altered gut motility of the stomach,
small bowel, or colon (irritable bowel or pseudoobstruction variants). There
is increasing evidence that upper GI motility disorders may masquerade as
pancreatobiliary-type pain. Multiple preliminary studies show disordered
duodenal motility in such patients. This area needs much more study to
determine the frequency, significance, and/or coexistence of these motor
disorders along with SOD.
Others have suggested that type III patients have duodenal-specific
visceral hyperalgesia with pain reproduction by duodenal distention, as well
as high levels of somatization, depression, obsessive–compulsive behav-
ior, and anxiety compared to control subjects. Patients with SOD appear
to have a higher than expected prevalence of the irritable bowel syndrome
(IBS), supporting the notion that SOD may occur as part of a more general-
ized functional disorder of the gut. Patients with SOD not only somatize
more than controls, but they may have an antecedent history of sexual or
0002086388.INDD 309 7/14/2014 4:59:44 PM

physical abuse similar to IBS. Selective SOD therapy cannot be expected to

provide symptom resolution in such patients and may account for the high
failure rate of sphincterotomy in many patients with type III SOD.
COMPLICATIONS AND THEIR PREVENTION

As noted earlier, patients with suspected SOD are at increased risk of post-
ERCP complications. While large multicenter studies and meta-analyses
have identified suspected SOD as an independent risk factor for post-
ERCP pancreatitis, several approaches may be undertaken to decrease the
incidence of this complication. Ensuring adequate pancreatic duct drain-
age post-ERCP with small diameter, temporary prophylactic pancreatic
stents in all patients with suspected SOD, whether or not sphincterotomy
is performed, may now be considered standard of care. As in all patients
undergoing ERCP, limiting the number of pancreatic duct injections and
extent of pancreatic duct opacification may further decrease pancreatitis
rates. Recently, rectal indomethacin was shown to reduce post-ERCP pan-
creatitis by nearly 50% in high-risk (mostly SOD) patients. Despite these
advances, complication rates remain significant in patients with suspected
SOD. These patients should be counseled thoroughly on the risk–benefit
ratio of undergoing ERCP with SOM, prior to proceeding with endoscopic
intervention. Figure 24.2 illustrates a suggested algorithm for the diagnos-
tic workup and treatment of patients with suspected type I, II, and III SOD.
Clinical history,
LFTs/amylase/lipase
Ultrasound/CT scan, EGD
+/− EUS, MRI/MRCP
Structural alterations No Structural alterations
Appropriate investigation Type I Type II Type III

and treatment or
Sphincterotomy Pharmacological trials
No response Response
Type II Type III
Further
investigation,
medical
therapy
Abnormal ERCP Normal

SOM with SOM SOM
Sphincterotomy Re-assess
FIGURE 24.2 Suggested algorithm for the diagnostic workup and treatment of patients
with suspected type I, II, or III SOD.
002086388.INDD 310 7/14/2014 4:59:45 PM

Suggested Readings
Behar J, Corazziari E, Guelrud M, et al. Functional gallbladder and sphincter of Oddi
disorders. Gastroenterology 2006;130:1498–1509.
Cheng CL, Sherman S, Watkins JL, et al. Risk factors for post-ERCP pancreatitis:
a prospective multicenter study. Am J Gastroenterol 2006;101:139–147.
Eversman D, Fogel EL, Rusche M, et al. Frequency of abnormal pancreatic and biliary
sphincter manometry compared with clinical suspicion of sphincter of Oddi dys-
function. Gastrointest Endosc 1999;50:637–641.
Freeman ML, Guda NM. Prevention of post-ERCP pancreatitis: a comprehensive
review. Gastrointest Endosc 2004;59:845–864.
Freeman ML, Gill M, Overby C, et al. Predictors of outcomes after biliary and pan-
creatic sphincterotomy for sphincter of Oddi dysfunction. J Clin Gastroenterol
2007;41:94–102.
Madura JA II, Madura JA. Diagnosis and management of sphincter of Oddi dysfunc-
tion and pancreas divisum. Surg Clin North Am 2007;87:1417–1429.
Morgan KA, Romagnuolo J, Adams DB. Transduodenal sphincteroplasty in the man-
agement of sphincter of Oddi dysfunction and pancreas divisum in the modern
era. J Am Coll Surg 2008;206(5):908–914; discussion 914–917.
Park SH, Watkins JL, Fogel EL, et al. Long-term outcome of endoscopic dual pan-
creatobiliary sphincterotomy in patients with manometry-documented sphinc-
ter of Oddi dysfunction and normal pancreatogram. Gastrointest Endosc
2003;57:483–491.
Sgouros SN, Pereira SP. Systematic review: sphincter of Oddi dysfunction—non-
invasive diagnostic methods and long-term outcome after endoscopic sphincter-
otomy. Aliment Pharmacol Ther 2006;24:237–246.
Venu RP, Geenen JE, Hogan WJ. Sphincter of Oddi stenosis and dysfunction. In: Sivak
MV Jr, ed. Gastroenterologic endoscopy, 2nd ed. Philadelphia, PA: WB Saunders,
2000:1023.
Dr. Fogel has provided an outstanding overview of sphincter of Oddi dys-

function, pathophysiology, and treatment. This editorial side bar is focused
on the technique of surgical transduodenal sphincteroplasty.
1. The patient is approached through upper midline or right subcostal inci-
sion. A wide Kocher maneuver is essential to mobilize the duodenum
into the operative field. Stay sutures placed on the duodenum are useful;
stay sutures placed adjacent to the papilla are helpful as well in elevating
the head of the pancreas and duodenum into the operative field; however,
care must be taken to avoid occluding the biliary or pancreatic sphincter.
The duodenum is opened longitudinally over the major papilla, which is
easily identifiable by palpation or ultrasound. In difficult cases, a probe
may be passed antegrade through the cystic duct and distal common bile
duct to identify the papillary orifice.
2. Once the papilla is exposed (see figures), the bile duct and pancre-
atic duct are intubated with small (5 French) pediatric feeding tubes.
Sphincteroplasty of the biliary sphincter is performed first. The tissue
may be divided sharply between fine-tipped clamps prior to suturing.
Alternatively, the sphincterotomy may simply be created/extended care-
fully using a needle tip cautery. If the latter approach is used, my prefer-
ence is to cauterize over the plastic feeding tube (not over a metal probe)
to minimize chance for inadvertent tissue injury. Care must be paid at
this point of the operation to avoid extending the sphincterotomy too far
(both in the biliary and pancreatic sphincter)—perforation into the ret-
roperitoneum is associated with potentially disastrous consequence. The
sphincteroplasty is completed by suturing the biliary epithelium to the
duodenal epithelium with long-acting absorbable sutures or nonabsorb-
able monofilament sutures. The pancreatic sphincteroplasty (septotomy)
0002086388.INDD 311 7/14/2014 4:59:45 PM

Operative (A) and schematic (B) illustration of sphincteroplasty with septoplasty.
is created between the biliary and pancreatic ductal orifice (see figures).
Again, care is taken here not to extend the septotomy too deeply. It is
prudent to leave a short segment (4 to 5 cm) of pediatric feeding tube in
the pancreatic duct to minimize chances of postoperative pancreatitis.
This tube is secured with chromic suture to ensure short-term stability
002086388.INDD 312 7/14/2014 4:59:47 PM

but promote early passage. Should the patient have pancreas divisum,
the minor papilla is addressed in a similar fashion.
3. The duodenotomy is closed transversely in two layers. Routine drainage
is not necessary; however, may be considered if concern for retroperito-
neal incursion is high.
4. Patients are typically counseled preoperatively to expect full liquid diet
or soft mechanical diet for the first few postoperative weeks. Immediate
postoperative laboratory evaluation includes liver chemistry and serum
amylase or lipase. Diet advancement is delayed if the patient manifests
biochemical pancreatitis. If patients have early postoperative epigastric
or right upper quadrant pain, abdominal x-ray should be obtained to
ensure passage of pancreatic stent.
0002086388.INDD 313 7/14/2014 4:59:47 PM

25 Common Bile Duct Stones
and Exploration
Attila Nakeeb
ETIOLOGY
Common bile duct (CBD) stones can be classified as either primary or
secondary. Primary CBD stones develop de novo within the bile ducts,
whereas secondary stones develop in the gallbladder and subsequently pass
into the CBD. In the United States, more than 85% of all bile duct stones are
secondary. Primary duct stones typically occur in conditions associated
with biliary stasis such as benign biliary strictures, sclerosing cholangitis,
choledochal cyst disease, oriental cholangiohepatitis, or sphincter of Oddi
dysfunction. Bile stasis promotes the overgrowth of bacteria in bile with
subsequent bilirubin deconjugation and the breakdown of biliary lipids,
resulting in the formation of brown pigment stones. Secondary bile duct
stones have a composition similar to gallbladder stones: approximately 75%
are cholesterol stones, and 25% are black pigment stones. The classifica-
tion of CBD stones into primary or secondary is of therapeutic importance
because primary stones require removal of the stones and a drainage pro-
cedure (choledochoenterostomy or sphincteroplasty), whereas secondary
stones can be treated by removal of the stones and cholecystectomy.
Most patients with CBD stones are asymptomatic. It is estimated
that CBD stones (choledocholithiasis) are found in approximately 10%
of patients with cholelithiasis. This incidence of CBD stones in patients
undergoing cholecystectomy ranges between 8% and 15%. The incidence
varies with age and is less than 5% in younger patients and more than 20%
in older patients with gallstones. Patients with symptomatic CBD stones
may present with biliary colic, extrahepatic biliary obstruction, cholangi-
tis, or pancreatitis. Typically, the pain and jaundice associated with CBD
stones are more intermittent and transient than when the biliary obstruc-
tion is caused by a malignancy. Fever and chills are present in patients with
cholangitis. Gallstone pancreatitis can develop from the obstruction of
the ampulla of Vater by common duct stones. In the west, CBD stones are
responsible for up to 50% of all cases of pancreatitis. Most patients with
gallstone pancreatitis experience a mild self-limited attack from which they
recover within a few days; however, some patients will progress to develop
severe pancreatitis with peripancreatic necrosis, infection, or other local
complications.
DIAGNOSIS
No single clinical variable is completely accurate in predicting the pres-
ence of choledocholithiasis. Therefore, the results of a detailed history and
physical examination, laboratory evaluations, and diagnostic imaging tests
must be considered together when assessing the likelihood that a patient
has CBD stones. Serum liver function tests (bilirubin, alkaline phospha-
tase, and transaminases) can be useful in predicting common duct stones.
314
002100650.INDD 314 7/14/2014 7:59:13 PM

Chapter 25 / Common Bile Duct Stones and Exploration 315
In the setting of suspected CBD stones, if any one value of the liver profile is
elevated, the risk for CBD stones approaches 20%. With two elevated values,
the risk increases to nearly 40%, and with three or more elevated values,
the risk for CBD stones is nearly 50%. Factors associated with a high risk of
choledocholithiasis include cholangitis, CBD size greater than 6 mm, CBD
stones visualized on ultrasound (US) or cross-sectional imaging, and jaun-
dice (bilirubin >4 mg/dL).
DIAGNOSTIC STUDIES
Transabdominal Ultrasonography
US is highly sensitive for the diagnosis of gallstones within the gallbladder.
Unfortunately, the sensitivity of transabdominal US for the detection of
CBD stones is only 30% to 50%. US successfully identifies the presence of
CBD stones in only 70% of patients because the distal bile duct is frequently
obscured by duodenal or colonic gas. On the other hand, transabdominal
US is very sensitive for identifying CBD dilation, which can suggest choledo-
cholithiasis. If the extrahepatic bile duct diameter is less than 3 mm, CBD
stones are exceedingly rare, whereas a CBD diameter greater than 10 mm in
a jaundiced patient is associated with CBD stones in more than 90% of cases.

Magnetic resonance cholangiopancreatography (MRCP) has a sensitivity of
90%, a specificity of 100%, and an overall diagnostic accuracy of 97% for the
diagnosis of CBD stones. MRCP (Fig. 25.1) permits direct imaging of the bili-
ary tract without the need for contrast, radiation exposure, or an invasive
FIGURE 25.1 MRCP demonstrating CBD stones.
0002100650.INDD 315 7/14/2014 7:59:14 PM

procedure. Disadvantages of MRCP include its relatively high cost, the need
for prolonged breath holding, and a lack of therapeutic options. MRCP can
be used to screen patients at low and moderate risk of having common duct
stones prior to endoscopic cholangiography. A normal magnetic resonance
cholangiogram can eliminate the need for an invasive endoscopic cholan-
giogram or intraoperative cholangiogram.
Endoscopic Ultrasound
Endoscopic ultrasound (EUS) is a semi-invasive test that can be performed
with a very low rate of complications (<0.1%). The sensitivity (92% to 100%)
and specificity (95% to 100%) for the diagnosis of CBD stones by EUS are
excellent. The negative predictive value for EUS is more than 97%. Therefore,
when EUS is negative for common duct stones, endoscopic retrograde chol-
angiopancreatography (ERCP) or intraoperative cholangiography (IOC)
can be avoided.
Direct Cholangiography
Direct cholangiography is the gold standard for diagnosing CBD stones.
Both endoscopic retrograde cholangiography (ERC) and percutaneous
transhepatic cholangiography (PTC) techniques can be used to directly
visualize the biliary tree. In addition to identifying CBD stones, endoscopic
cholangiography (Fig. 25.2) permits safe removal of most stones and is there-
fore the preferred approach for patients with suspected CBD stones. Skilled
endoscopists can successfully cannulate the CBD in 90% to 95% of patients.
Complications of diagnostic cholangiography including pancreatitis and
cholangitis occur in approximately 5% of patients; the vast majority of
these complications are mild and self-limiting. ERC may be unsuccessful in
patients with previous gastric surgery (Billroth II reconstruction, Roux-en-Y
FIGURE 25.2 ERCP demonstrating CBD stones.
002100650.INDD 316 7/14/2014 7:59:15 PM

gastric bypass), periampullary diverticula, or tortuous bile ducts. PTC may

be used to image the bile ducts if ERC is unsuccessful.
Intraoperative Cholangiography
IOC also can be successfully accomplished in more than 95% of cases. The
cholangiogram should be carefully evaluated for filling defects within the
ducts, presence of contrast in the duodenum, and the intrahepatic bili-
ary anatomy (Fig. 25.3). Importantly, the surgeon must be able to interpret
IOC images in real time, in the operating room for this study to be valu-
able. Debate continues over the need to perform routine IOC at the time
of cholecystectomy. Advocates of routine IOC argue that asymptomatic
CBD stones can be identified and biliary injuries prevented by performing
routine IOC. Critics of this approach suggest that the incidence of retained
stones is no greater when cholangiography is performed selectively based
on clinical and laboratory criteria. The indications for performing cholangi-
ography during cholecystectomy include (a) a dilated CBD, (b) a wide cystic
duct, (c) palpable CBD stones, (d) elevated serum liver function tests or bili-
rubin, and (e) a history of pancreatitis. If these criteria are strictly followed,
approximately 30% of patients will require IOC at the time of cholecystec-
tomy. In addition to defining biliary anatomy, IOC can identify the size,
number, and location of CBD stones. This information is critical in choosing
the most appropriate treatment for CBD stones.
Intraoperative Ultrasonography
Intraoperative ultrasonography also can be used to identify CBD stones at the
time of cholecystectomy. In experienced hands, intraoperative ultrasonogra-
phy has been shown to be comparable to IOC for the diagnosis of CBD stones.
Laparoscopic ultrasonography is performed with a high-frequency (7.5- to
10-mHz) probe, and the bile duct is imaged in the transverse and longitudinal
planes. The distal bile duct can be visualized in more than 95% of cases.
Right hepatic Left hepatic

duct duct
Cystic duct
Dilated common
bile duct
Meniscus of stone
blocking CBD
FIGURE 25.3 Intraoperative cholangiogram. (From

Lillemoe K, Jarnigan W. Master techniques in surgery:
hepatobiliary and pancreatic surgery. Philadelphia, PA:
Lippincott Williams & Wilkins, 2012.)
0002100650.INDD 317 7/14/2014 7:59:16 PM

General Diagnostic Strategy

The choice of radiologic studies used to evaluate a patient with suspected
choledocholithiasis should be based on the probability of this diagnosis.
A simple algorithm for the workup of patients with CBD stones is shown in
Figure 25.4. Patients at highest risk for choledocholithiasis should undergo
endoscopic cholangiography. Patients at intermediate risk may be screened
with magnetic resonance cholangiography or EUS and proceed to laparo-
scopic cholecystectomy if no stones are identified. Those patients at low
risk of harboring common duct stones may be evaluated with IOC at the
time of laparoscopic cholecystectomy with laparoscopic CBD exploration
or postoperative endoscopic stone extraction reserved for the few patients
with a positive study.
MANAGEMENT
Currently, several options are available to the surgeon for the treatment of
CBD stones. In choosing the most appropriate approach for an individual
patient, factors such as the local endoscopic expertise, the surgeon’s laparo-
scopic skill, and the patient’s clinical condition must be considered.
ENDOSCOPIC THERAPY
ERC with endoscopic sphincterotomy permits CBD stones to be removed
without the need for conventional surgery. Stones can be successfully
removed from the CBD in 85% to 95% of cases. The endoscopic approach
is particularly useful for patients prior to cholecystectomy in whom a high
suspicion exists for CBD calculi, particularly if laparoscopic CBD explora-
tion is not available. Endoscopic clearance of stones from the CBD precho-
lecystectomy can obviate the need for an open operation. Furthermore, if
Symptomatic cholelithiasis +/− choledocholithiasis
Debilitated Jaundice
Elderly Dilated CBD
Severe Pancreatitis
Cholangitis
ERC / ES Elevated LFTs Contraindications
to Laparoscopy
Follow ERC / ES
Laparoscopic Open
Cholecystectomy Cholecystectomy
Intraoperative
Retained Stones Cholangiogram Stones
Stones or Ultrasound
Transcystic
ERC / ES Extraction Open CBDE Drainage
T-tube Procedure
Stones
Postoperative CBDE / t-tube Stones No Stones

ERC / ES
Remove
Stones Removal Via T-tube
T-tube
FIGURE 25.4 Algorithm for the management of CBD stones. (From Mulholland MW,
Lillemoe KD, Doherty GM, et al. Greenfield’s surgery: scientific principles & practice, 5th
ed. Philadelphia, PA: Lippincott Williams & Wilkins, 2010.)
002100650.INDD 318 7/14/2014 7:59:17 PM

endoscopic stone extraction is not possible because of multiple gallstones,

intrahepatic stones, large gallstones, impacted stones, duodenal diver-
ticula, prior gastrectomy, or bile duct stricture, this information is known
preoperatively, and an open CBD exploration or drainage procedure can
be performed at the time of cholecystectomy. After sphincterotomy, most
stones smaller than 1 cm in diameter pass spontaneously. A balloon cath-
eter or stone basket also can be used to retrieve stones if needed. If endo-
scopic clearance is incomplete, an endoscopic stent can be placed into the
CBD to maintain drainage and prevent cholangitis.
Endoscopic sphincterotomy and stone extraction is well tolerated in
most patients. Complications occur in 5% to 8% of patients and include
cholangitis, pancreatitis, perforation, and bleeding. The overall mortal-
ity rate is 0.2% to 0.5%. Complete clearance of all common duct stones is
achieved endoscopically in 71% to 75% of patients at the first procedure and
in 84% to 95% of patients after multiple endoscopic procedures.
Preoperative ERC plus endoscopic sphincterotomy is the preferred
management option for CBD stones in several conditions. In the setting
of acute suppurative cholangitis, morbidity and mortality are significantly
decreased if preoperative biliary decompression and stone removal are
accomplished before cholecystectomy. Patients with severe gallstone pan-
creatitis or a significant deterioration of their clinical condition also have
been shown to benefit from early ERC and stone clearance. Cholecystectomy
can then be performed after the pancreatitis has resolved. In patients with a
dilated CBD (>8 mm) on US and jaundice, an ERC also should be performed
to rule out a malignancy or biliary stricture that would alter the surgical
management. In patients with a high operative risk, ERC plus endoscopic
sphincterotomy can be performed or definitive treatment to remove CBD
stones and the gallbladder left in place.
LAPAROSCOPIC THERAPY
Laparoscopic exploration of the CBD for choledocholithiasis enables
appropriately selected patients to undergo complete management of
their calculus biliary tract disease with one procedure. The laparoscopic
approach is ideal for patients with CBD stones identified during IOC or US.
The two approaches for laparoscopic bile duct exploration are laparoscopic
transcystic duct bile duct exploration (LTCBDE) or laparoscopic choledo-
chotomy (Table 25.1). The indications for transcystic duct exploration are
TABLE
Laparoscopic Transcystic CBD Exploration versus
25.1 Laparoscopic Choledochotomy
Transcystic Choledochotomy
Stones
Number <8 Any
Size <9 mm Any
Location Distal to cystic duct Entire duct
Bile duct size Any >6 mm
Drain Optional cystic duct tube T-tube
Contraindications Friable cystic duct Small diameter duct
Intrahepatic stones Inability to suture
laparoscopically
Advantages Multiple large stones T-tube for postoperative
No T-tube access
Short hospital stay
0002100650.INDD 319 7/14/2014 7:59:17 PM

filling defects noted on cholangiography (CBD stones), stones less than 9

mm in diameter, stones distal to the cystic duct entrance to the bile duct,
and less than six stones. Contraindications to LTCBDE are a small friable
cystic duct, more than eight stones in the CBD, common hepatic duct
stones, and stones larger than 1 cm. Laparoscopic choledochotomy can
be performed if LTCBDE fails or is contraindicated, if stones are present
proximal to the cystic duct, or when multiple stones are present. The only
contraindication to laparoscopic choledochotomy is a small common duct
(<6 mm) that might be narrowed during its closure.
Laparoscopic Transcystic Duct Bile Duct Exploration

The technique of LTCBDE involves careful dissection of the cystic duct
down to its junction with the CBD. A clip is placed on the gallbladder side
of the cystic duct and a ductotomy is created sharply. After completing a
cholangiogram, a guide wire should be inserted into the bile duct. A chol-
angiocatheter can then be advanced over the guide wire into the bile duct
and saline irrigated through the catheter in an attempt to flush small stones
out of the bile duct. Glucagon should be administered intravenously to
encourage relaxation of the sphincter of Oddi. If the stones in the CBD are
larger than the lumen of the cystic duct, the cystic duct can be dilated with
a series of dilators using either the Seldinger technique or an angioplasty
balloon catheter. Stone retrieval baskets can be inserted over the guide wire
and stones extracted under fluoroscopic guidance. Flexible choledochos-
copy via the cystic duct also can be performed and stones extracted under
direct vision using baskets (Fig. 25.5). Success rates greater than 95% have
been reported for bile duct clearance using choledochoscopy. To document
FIGURE 25.5 Laparoscopic transcystic duct CBD exploration. (From Lillemoe K,

Jarnigan W. Master techniques in surgery: hepatobiliary and pancreatic
002100650.INDD 320 7/14/2014 7:59:18 PM

stone clearance, a completion cholangiogram should be performed. A cystic

duct drainage tube can be left in place if findings on the cholangiogram are
equivocal. This tube can be used postoperatively for cholangiography and
radiographic treatment of retained stones if necessary. After LTCBDE, the
cystic duct stump should be secured with an endoloop or a suture ligature
and not be clipped.
Laparoscopic Choledochotomy
Laparoscopic choledochotomy (Fig. 25.6) is another excellent approach to
CBD stones when the CBD diameter is 6 mm or greater. The anterior wall
of the CBD is bluntly dissected and a longitudinal choledochotomy created
in the anterior wall of the bile duct distal to the cystic duct insertion. The
choledochotomy should be fashioned to accommodate the diameter of the
largest stone. Two stay sutures can be placed in the common duct and used
to tent up the anterior CBD wall to facilitate incision. A larger choledocho-
scope (3.3 mm, 2.4-mm working channel) can then be placed into the bile
duct and stones extracted with baskets or balloon catheters. The choledo-
chotomy is then closed over a T-tube with a 4-0 absorbable suture.
Advantages of laparoscopic choledochotomy relative to LTCBDE
include the ability to remove larger stones (>1 cm), to remove stones from
the proximal hepatic ducts, to remove multiple stones, and to use biliary
lithotripsy to fragment impacted stones. The disadvantages of laparoscopic
choledochotomy are that it requires a T-tube and considerable laparo-
scopic suturing skill to close the choledochotomy.
Clearance of all CBD stones is achieved in 75% to 95% of patients
with laparoscopic CBD exploration. Results from several randomized
FIGURE 25.6 Laparoscopic choledochotomy. (From Lillemoe K, Jarnigan W.

Master techniques in surgery: hepatobiliary and pancreatic surgery.
0002100650.INDD 321 7/14/2014 7:59:20 PM

rospective trials comparing a single-stage approach (laparoscopic

p
cholecystectomy + laparoscopic common bile duct exploration) versus a
two-stage approach (laparoscopic cholecystectomy + either pre- or post-
operative ERCP) are shown in Table 25.2. The morbidity and mortality of
laparoscopic CBD exploration are similar to laparoscopic cholecystectomy
alone. The overall hospital length of stay tends to be shorter for patients
managed with a one-stage approach versus a two-stage approach.
Open Common Bile Duct Exploration

Open CBD exploration is performed much less frequently with the
increased use of endoscopic, percutaneous, and laparoscopic techniques
to remove CBD stones. The first step is to perform a full Kocher maneuver
mobilizing the duodenum so that a hand can be placed behind the head
of the pancreas and the distal CBD palpated. Any impacted stones may be
milked proximally. The supraduodenal bile duct is then exposed, and two
stay sutures are placed in the CBD just below the cystic duct. The anterior
wall of the bile duct is then elevated with the stay sutures and a longitudi-
nal choledochotomy made. The bile duct then can be explored for stones.
Rigid instruments should not be used to extract stones because they can
injure the delicate ductal epithelium. A soft rubber irrigating catheter can
be used to gently flush out any stones or debris. Balloon-tipped catheters
then can be passed proximally and distally into the ducts to retrieve stones.
Adequate clearance of the duct should be confirmed visually with flexible
choledochoscopy. Remaining stones can be removed by irrigation or by the
use of instruments such as stone forceps, wire baskets, or balloon catheters.
A T-tube should be placed in the bile duct and the choledochotomy closed
with 4-0 absorbable suture. Completion cholangiography is performed
before closing the abdomen to rule out the presence of retained stones or a
bile leak around the T-tube. Postoperatively, a T-tube cholangiogram is per-
formed 3 to 7 days after the exploration. If the cholangiogram is normal, the
tube can be clamped and the tube pulled 3 to 6 weeks later. If retained stones
are detected, the tract is allowed to mature, and percutaneous extraction
can be performed by a radiologist in 4 to 6 weeks. Open CBD exploration
can be accomplished with almost no mortality in patients younger than age
60 years but carries mortality of up to 4% in septic patients.
TABLE
Select Randomized Trials of One- versus Two-Stage
25.2 Management of CBD Stones
Duct Length
Clearance Morbidity of Stay
Author Year Treatment N (%) (%) (days)
Rhodes 1998 LC + ERCP 40 93 15 3.5
LC + LCBDE 40 75 5 1
Cuschieri 1999 ERCP + LC 133 62 13 9
LC + LCBDE 133 70 16 6
Nathanson 2005 LC + ERCP 45 96 24 7.7
LC + LCBDE 41 97 29 6.4
Noble 2009 ERCP + LC 47 62 34 3
LC + LCBDE 44 86 52 5
Rogers 2010 ERCP + LC 55 98 9 5
LC + LCBDE 57 88 11 4
LC, laparoscopic cholecystectomy; ERCP, endoscopic retrograde cholangiopancreatography; LCBDE,
laparoscopic common bile duct exploration.
002100650.INDD 322 7/14/2014 7:59:20 PM

DRAINAGE PROCEDURES
Patients with an impacted CBD stone at the ampulla that cannot be
removed with a CBD exploration or those with multiple stones in a non-
dilated duct may require a transduodenal sphincteroplasty. A sphinc-
teroplasty also is indicated in the presence of an ampullary stenosis or a
choledochocele. The first step in a sphincteroplasty is to perform a Kocher
maneuver. A small longitudinal duodenotomy is made over the ampulla,
and two stay sutures are placed on each side of the ampulla to elevate it.
A small incision is made at the 11 o’clock position in the sphincter taking
care to avoid the pancreatic duct, which is usually found at the 5 o’clock
position. The sphincterotomy is extended through the sphincter (approxi-
mately 1.5 cm) and the impacted stone removed. The bile duct and duo-
denal mucosa are then approximated with interrupted 4-0 absorbable
sutures. The duodenotomy should be closed transversely to prevent nar-
rowing of the duodenal lumen.
Patients with grossly dilated bile ducts (>2 cm), multiple stones
(>5), intrahepatic stones, primary CBD stones, or a distal biliary stricture
should be considered for a biliary drainage procedure. The two options
are a choledochoduodenostomy and a Roux-en-Y choledochojejunostomy.
Choledochoduodenostomy can be performed in either a side-to-side or an
end-to-side fashion. Advantages of a choledochoduodenostomy are that
it can be performed rapidly, it requires only one anastomosis, and the bile
duct still can be accessed endoscopically. However, a side-to-side anasto-
mosis leaves the distal CBD in continuity and can lead to the “sump syn-
drome.” In this situation, food debris from the duodenum can enter the
distal limb of the CBD and obstruct the anastomosis or the pancreatic duct
orifice leading to cholangitis or pancreatitis. Long-term complications of
choledochoduodenostomy include a small but real incidence of cholangitis
and liver abscess.
Roux-en-Y hepatico- or choledochojejunostomy also is an excellent
option for biliary drainage. This operation is performed by dividing the
hepatic or common duct and doing an end-to-side anastomosis to a 60-cm
Roux limb of jejunum. Development of the sump syndrome is not a concern
because a hepatico- or choledochojejunostomy completely diverts bile flow
from the alimentary stream.
CONCLUSIONS
CBD stones may be approached endoscopically, laparoscopically, or by
open CBDE. Choice of therapy depends on both patient considerations as
well as local endoscopic and surgical expertise. Patients with primary CBD
stones must be considered for a biliary drainage procedure.
Suggested Readings
Abboud BA, Malet PF, Berlin JA, et al. Predictors of common bile duct stones prior to
cholecystectomy: a meta-analysis. Gastrointest Endoscopy 1996;44:450–459.
Alexakis N, Connor S. Meta-analysis of one vs. two-stage laparoscopic/endoscopic
management of common bile duct stones. HPB 2012;14:254–259.
Duncan C, Riall T. Evidence-based surgical practice: calculous gallbladder disease.
J Gastrointest Surg 2012;16:2011–2025.
Freeman M, Nelson D, Sherman S, et al. Complications of endoscopic biliary sphinc-
terotomy. N Eng J Med 1996;335:909–918.
Kharbutli B, Velanovich V. Management of preoperatively suspected choledocholithia-
sis: a decision analysis. J Gastrointest Surg 2008;12:1973–1980.
Lillemoe K, Jarnigan W. Master techniques in surgery: hepatobiliary and pancreatic
surgery. Philadelphia, PA: Lippincott Williams & Wilkins, 2012.
0002100650.INDD 323 7/14/2014 7:59:20 PM

Liu T, Consorti E, Kawashima A, et al. Patient evaluation and management with

selective use of magnetic resonance cholangiography and endoscopic retro-
grade cholangiopancreatography before laparoscopic cholecystectomy. Ann Surg
2001;234:33–40.
Mulholland MW, Lillemoe KD, Doherty GM, et al. Greenfield’s surgery: scientific prin-
ciples & practice, 5th ed. Philadelphia, PA: Lippincott Williams & Wilkins, 2010.
Rhodes M, Sussman L, Cohen L, et al. Randomized trial of laparoscopic exploration of
common bile duct versus postoperative endoscopic retrograde cholangiography
for common bile duct stones. Lancet 1998;351:159–161.
Vilgrain V, Palazzo L. Choledocholithiasis: role of US and endoscopic ultrasound.
Abdom Imaging 2001;26:7–14.
002100650.INDD 324 7/14/2014 7:59:20 PM

26 Choledochal Cyst
Alessandro Paniccia and Barish H. Edil
EPIDEMIOLOGY
Choledochal cysts (CCs) are more appropriately defined as a cystic disorder
of the biliary ducts. This rare condition is characterized by dilatation of the
extrahepatic and/or intrahepatic biliary tree. It was originally described as
a disease of infancy, but in recent years, an increase number of cases have
been diagnosed during adulthood.
The prevalence of CCs is subject to significant geographic variation.
The highest incidence, 1:1,000 births, is observed in Southeast Asian coun-
tries. The incidence of CCs in the western world is between 1:100,000 and
1:150,000 and is described within the US population as 1:13,500 live births.
In general, the female-to-male ratio of CC is approximately 4:1.
The diagnosis of cystic disorder of the biliary ducts carries with it a
significant lifetime risk of malignancy that appears to increase with age.
The overall risk of malignant degeneration is approximately 10%.
ETIOLOGY
Different theories have been presented to describe the etiology of cystic dila-
tation of the biliary ducts. These theories postulate that anatomical and func-
tional abnormalities of the biliary ducts combined with increased intraductal
pressure and pancreatic enzyme activation contribute to bile duct dilatation.
Babbitt et al. in 1969 popularized the “long common channel” theory
recognizing an abnormal junction of the main pancreatic and common bile
ducts (APBDJ) outside the duodenal wall and proximal to the sphincter of
Oddi. This anatomic anomaly may allow pancreatic secretion to reflux into
the biliary tree. The subsequent enzyme activation leads to chronic inflam-
mation of bile duct epithelium, with ductal wall injury and ultimately cystic
dilation. Although APBDJ is a well-recognized anatomic anomaly and most
accepted hypothesis, it is only present in 80% of all CCs.
CLASSIFICATION
The first classification system for CC was proposed by Alonso-Lej et al. in
1959 and described three different types of cystic dilation. The Todani modi-
fication of the Alonso-Lej classification was introduced in 1977 and is cur-
rently used. Todani expanded the original classification system to include
intrahepatic disease resulting in a total of five types of cystic disorders
based on site, shape, and extent of biliary tree involvement (Fig. 26.1).
Type I cyst is a dilation of the common hepatic duct and represents
50% to 80% of all cysts. Type 1 CC is subclassified into three types based on
shape: type Ia cystic, type Ib focal, and type Ic fusiform.
Type II cyst is a supraduodenal diverticulum of the extrahepatic duct
to which it is connected with a narrow stalk. It represents only 2% to 3% of
all cysts.
325
0002086393.INDD 325 7/14/2014 8:04:06 PM

FIGURE 26.1 Classification of biliary cysts.

A. Type I. B. Type II. C. Type III. D. Type IV.
E. Type V. (From Lipsett PA, Pitt HA,
Colombani PM, et al. Choledochal disease:
a changing pattern of presentation. Ann
Surg 1994;220(5):644.)
002086393.INDD 326 7/14/2014 8:04:08 PM

Chapter 26 / Choledochal Cyst 327
Type III cyst, also known as choledochocele, is a dilation of the com-

mon bile duct confined to the wall of the duodenum. Type III CC represents
less than 10% of all cysts.
Type IV cyst refers to multiple cystic dilations of the biliary tract and
is further subclassified into type IVa where multiple cysts are present in the
intrahepatic and extrahepatic bile ducts and type IVb where multiple cysts
only involve the extrahepatic bile duct. Type IVa cyst is the second most
common type of CC representing 30% to 40%, while type IVb only accounts
for less than 5% of all cysts.
Type V cyst (Caroli disease) involves only the intrahepatic biliary
ducts and can be unilobar or bilobar. It represents less than 10% of all CCs.
DIAGNOSIS
Presentation
The historically described pathognomonic triad of abdominal pain, jaun-
dice, and right upper quadrant palpable mass is only present in 10% to 17%
of children. However, 85% of children present with at least two symptoms of
the triad compared to only 25% of adults.
Vague right upper quadrant abdominal pain associated with nausea
and emesis is often the presenting symptom in adults. Laboratory evalua-
tion of liver function and pancreatic enzymes frequently reflects common
conditions such as cholangitis or pancreatitis, where concurrent pancreati-
tis may be present in 30% to 70% of adult patients. Additionally, concurrent
cystolithiasis and cholecystolithiasis are present in approximately 70% of
patients. As a result, an interval of up to 6 years from the first presenting
symptoms to the diagnosis of cystic disorder of the biliary ducts is relatively
common. This delay in diagnosis often leads to additional procedures such
as cholecystectomy or other surgical exploration. Most often, adults are
diagnosed with CC during evaluation and treatment of complications such
as cholecystitis or acute pancreatitis. Diagnosis may also be discovered
incidentally during radiologic investigation for other problems. In contrast
to other cyst types, patients with type III choledochal cyst (choledochocele)
commonly present with acute pancreatitis.
Although rare, hepaticolithiasis and intrahepatic abscess have been
described as presenting complications of type IV and V bile duct cysts. This
can eventually lead to secondary biliary cirrhosis and portal hypertension
with subsequent liver failure.
Cyst rupture and biliary peritonitis have been also described as pre-
senting symptoms but usually in children. These situations are extraordi-
narily rare.
The most common malignancy associated with CC disease is cholan-
giocarcinoma, though the presence of CC also increases the risk of developing
gallbladder cancer.
Imaging
Demonstration of the continuity of the cystic lesion with the biliary duct
is paramount in distinguishing it from other intra-abdominal cysts such
as pancreatic pseudocyst, biliary cystadenoma, or echinococcal cysts. In
patients with equivocal bile duct dilation (i.e., 1 to 2 cm), the presence of
APBDJ strongly suggests the diagnosis of CC.
Abdominal ultrasound (US) is often the primary and the most cost-
effective imaging modality in diagnosing bile duct cyst; the sensitivity of US
ranges from 71% to 97%.
Endoscopic ultrasound has been described in the literature in order
to overcome the intrinsic limitations of transabdominal US such as body
habitus, bowel gas, the presence of overlying intra-abdominal structures,
0002086393.INDD 327 7/14/2014 8:04:08 PM

and poor visualization of distal bile duct. It also gives you information of the
pancreas and biliary duct anatomy.
Computed tomography (CT) scan is now widely available and pres-
ents several advantages in the diagnosis of CC. Some of these advantages
include characterization of intrahepatic bile ducts, dilation evaluating the
cysts surrounding structures, and more visualization of the distal bile duct
and the pancreatic head. It also allows for evaluation of malignancy that
may mimic CC or may be originating from a CC. (i.e., very thick cyst wall or
portal lymphadenopathy).
Computed tomography cholangiography (CTCP) though not widely
available currently has 90% sensitivity for diagnosis of CC and can correctly
visualize the bile duct anatomy prior to a surgical procedure.
Magnetic resonance cholangiography (MRCP) is becoming the gold
standard in diagnosis with a reported sensitivity ranging between 90% and
100%. Though MRI/MRCP is relatively more expensive than CT, MRCP offers
similar excellent cross-sectional imaging evaluation as well as superior
visualization of the entire biliary tree (including APBDJ).
Particularly challenging in children is proper distinction between
biliary atresia and bile duct cyst. In this situation, a 99technetium HIDA
scan will demonstrate continuity of the cystic lesion with the bile duct and
eventually emptying of contrast into the bowel. Retention of contrast in the
biliary system is typical of biliary atresia or any condition causing distal
common bile duct (CBD) obstruction. The sensitivity of a hepatobiliary imi-
nodiacetic acid (HIDA) scan, however, is subject to wide variation ranging
from 100% for type I to only 67% for type IVa. This inaccuracy is mainly due
to HIDA inability to visualize the intrahepatic bile ducts.
In light of the increased accuracy of MRCP and CTCP, invasive chol-
angiography modalities such as percutaneous transhepatic cholangiogram
(PTC) and endoscopic retrograde cholangiopancreatography are applied
less frequently. However, in the scenario where biliary drainage is inade-
quate (type IV/V), PTC and stent placement may be required.
An exception is made for type III CC in which endoscopy is considered
the principal diagnostic modality because of its potential to also provide
therapeutic intervention. This modality allows cannulation of the papilla of
Vater and eventually opacification of the dilated intramural common bile
duct. In addition, sphincterotomy is therapeutic for these cysts.
MANAGEMENT
Surgical treatment of CCs has undergone significant changes during the
last century. Several surgical techniques have been described including cyst
marsupialization, choledochorrhaphy, and internal drainage via cystenter-
ostomy to the duodenum or jejunum. The results with the above techniques
have been unsatisfying due to significant associated morbidity and mortal-
ity such as ascending cholangitis, anastomosis stricture, stone formation,
and a remarkable 30% incidence of postoperative malignancy occurring
15 years earlier than primary malignancy. Therefore, any treatment after
complete resection is of historical interest only.
The general consensus today is that surgical management depends on
cyst type and that complete cyst excision should be obtained in order to
decrease the risk of associated malignancy.
Type I CC is successfully treated by complete cyst resection with a
Roux-en-Y hepaticojejunostomy either in an open fashion or, when techni-
cally feasible, with a laparoscopic approach.
The technical aspects of this operation involve mobilization of the
hepatic flexure and wide Kocher maneuver because the cyst extends
002086393.INDD 328 7/14/2014 8:04:08 PM

osterior to the duodenum into the head of the pancreas. Dissection on the
p
anterior cyst is continued caudally into the pancreatic head until the bile
duct narrows, which is the inferior portion of the cyst. This area is ligated,
and then, the cyst is reflected cephalad. Dividing the bile duct and displacing
the cyst anteriorly allow identification of the portal vein. Posterior dissection
is continued until the bile duct is transected just distal to the hepatic duct
confluence. Care should be taken not to injure the right hepatic artery, which
commonly passes posterior to the bile duct and anterior to the portal vein.
Depending on bile duct size, some biliary surgeons prefer to reconstruct this
anastomosis across soft Silastic transhepatic stents. A 60-cm limb is brought
up in a retrocolic fashion for a Roux-en-Y hepaticojejunostomy. There is an
associated 7% complication rate including early anastomotic leak, pancre-
atic leak due to injury to the pancreatic duct, and bowel obstruction. Late
complications include cholangitis, pancreatitis, and anastomotic stricture.
The important point is that complete cyst excision is required to prevent
future degeneration of CC into cholangiocarcinoma.
Type II CC can be treated with simple cyst excision. The procedure
of choice depends on cyst size and location. Commonly used surgical
approaches are an extrahepatic biliary resection and Roux-en-Y reconstruc-
tion or complete excision with primary closure over a T tube. In order to
prevent segmental bile duct narrowing, the defect in the wall of the common
bile duct should be closed transversely or over a T tube.
Type III CCs have an extremely low risk of malignancy. These cysts
more commonly present with pancreatitis or abdominal pain. They may
also be diagnosed in the setting of pancreatic or biliary obstructive symp-
toms or much more rarely gastric outlet syndrome. Different treatments
have been proposed ranging from simple endoscopic sphincterotomy to
complete cyst excision through a lateral duodenostomy. If operation is
undertaken, it is of primary importance to identify the pancreatic duct in
order to avoid accidental injuries. Cannulation of the pancreatic duct with
Silastic tubes prior to cyst resection can be used to identify the pancreatic
duct and avoid injuries. Eventually the cyst can be excised and sphincter-
oplasty can be performed. Some authors believe that the pancreatic and
biliary duct should be routinely separated and reanastomosed to the duo-
denum in order to avoid pancreaticobiliary mixing. Most authorities feel
that endoscopic “unroofing” provides adequate therapy for type III CC.
Type IV cyst treatment is dictated by the presence or absence of intra-
hepatic involvement. When intrahepatic involvement is diffuse, complete
cyst removal requires partial hepatectomy.
Type IVb is treated in the same fashion of type I cyst.
The treatment for type IVa cyst depends on the extent of the intra-
hepatic involvement and the presence of cirrhosis and/or portal hyperten-
sion. If the intrahepatic involvement is confined to a single lobe, a wide hilar
hepaticojejunostomy with segmental hepatectomy can be performed. If the
intrahepatic component is diffuse and bilobar involvement is encountered,
extrahepatic biliary excision should still be performed, and drainage of
intrahepatic biliary tree should be considered with two large-bore Silastic
stents. The aim of this approach is to reduce intrahepatic biliary stasis, chol-
angitis, and stone formation potentially decreasing the risk of liver damage
and malignant transformation. Some authors recommend having trans-
plantation in this setting; treatment of patients with type IV CCs must be
determined in the setting of an experienced multidisciplinary group.
Type V cyst (Caroli disease) treatment is dictated by extent of intrahe-
patic disease and the presence or absence of cirrhosis, portal hypertension,
and liver failure. Early recognition of this disease is paramount in order to
avoid recurrent cholangitis, biliary stasis, and intrahepatic abscess formation.
0002086393.INDD 329 7/14/2014 8:04:08 PM

Initial nonoperative management may be warranted to treat cholan-

gitis, intrahepatic stone formation, as well as biliary stasis in the attempt to
decrease progressive liver injury.
In the absence of portal hypertension, if there is diffuse liver disease
extrahepatic bile duct excision, biliary enteric anastomosis and permanent
drainage of the intrahepatic cyst with large-bore Silastic stents may be con-
sidered. In cases of unilobar liver involvement, the treatment consists of
extrahepatic bile duct excision and segmental hepatectomy. Orthotropic
liver transplant is an option for Caroli disease with bilobar cystic involve-
ment. Outcomes of transplantation are comparable to those for other trans-
plant populations; however, the timing of transplant remains controversial.
OUTCOME/FOLLOW-UP
The treatment of choledocal cyst disease with complete resection is suc-
cessful in preventing degeneration into cancer. Early complications include
biliary anastomotic leak, pancreatic leak in case of accidental pancreatic
duct injury, and bowel obstruction due to intussusception or internal
hernia.
Late complications include biliary anastomotic stricture and recur-
rent cholangitis. These complications are seen in approximately 25% to 35%
of the cases and are often associated with stone formation. Percutaneous
dilation of anastomotic stricture with biliary stent placement usually pro-
vides durable treatment.
Despite excision of CCs, the risk of malignancy in the remaining bile
ducts is estimated between 0.7% and 6% likely due to subclinical malignant
disease not detected at the time of surgery or incomplete cyst excision.
Therefore, patients diagnosed with cystic disorder of the biliary ducts require
lifelong postoperative surveillance with serial imaging and monitoring of
serum liver enzymes and tumor markers (CA 19–9).
CONCLUSION
Although cystic disorder of the biliary ducts remains a rare condition, in
recent years, it is being recognized with increased frequency in the western
population.
The importance of CCs is related to the lifetime risk of malignancy of
approximately 10%. Cholangiocarcinoma remains the most common type
of cancer arising from CCs, though these patients also have an increased
risk for developing gallbladder cancer. Complete surgical resection of dis-
eased bile ducts is the treatment of choice and should be pursued when
technically possible in reasonable surgical candidates. When surgical
excision is not possible, interventions should be considered such as tran-
shepatic drains in order to decrease biliary stasis and improve symptom-
atology. Complete excision of the cyst can prevent biliary cancer. If CC
disease is left behind as seen with type IV and type V, lifelong management
and cancer surveillance are necessary. Even with complete CC excision,
lifetime surveillance should continue, as late complications are common.
Suggested Readings
Abbas HM, Yassin NA, Ammori BJ. Laparoscopic resection of type I choledochal cyst
in an adult and Roux-en-Y hepaticojejunostomy: a case report and literature
review. Surg Laparosc Endosc Percutan Tech 2006;16(6):439–444.
Alonso-Lej F, Rever Wb, Pessagno Dj. Congenital choledochal cyst, with a report of 2,
and an analysis of 94, cases. Int Abstr Surg 1959;108:1–30.
Ammori JB, Mulholland MW. Adult type I choledochal cyst resection. J Gastrointest
Surg 2009;13(2):363–367.
002086393.INDD 330 7/14/2014 8:04:08 PM

Babbitt DP. Congenital choledochal cysts: new etiological concept based on anoma-
lous relationships of the common bile duct and pancreatic bulb. Ann Radiol
1969;12:231–240.
Dhupar R, Gulack B, Geller DA, et al. The changing presentation of choledochal cyst
disease: an incidental diagnosis. HPB Surg 2009;2009:103739.
Edil BH, Cameron JL, Reddy S, et al. Choledochal cyst disease in children and adults:
a 30-year single-institution experience. J Am Coll Surg 2008;206(5):1000–1005;
discussion 1005–1008.
Edil BH, Olino K, Cameron JL. The current management of choledochal cysts. Adv
Surg 2009;43:221–232.
Lipsett PA, Pitt HA. Surgical treatment of choledochal cysts. J Hepatobiliary Pancreat
Surg 2003;10(5):352–359.
Metcalfe MS, Wemyss-Holden SA, Maddern GJ. Management dilemmas with chole-
dochal cysts. Arch Surg 2003, 138(3): 333–339.
Singham J, Yoshida EM, Scudamore CH. Choledochal cysts: part 1 of 3: classification
and pathogenesis. Can J Surg 2009;52(5):434–440.
Singham J, Yoshida EM, Scudamore CH. Choledochal cysts: part 2 of 3: Diagnosis. Can
J Surg 2009;52(6):506–511.
Singham J, Yoshida EM, Scudamore CH. Choledochal cysts. Part 3 of 3: management.
Can J Surg 2010;53(1):51–56.
Todani T, Watanabe Y, Narusue M, et al. Congenital bile duct cysts: classification,
operative procedures, and review of thirty-seven cases including cancer arising
from choledochal cyst. Am J Surg 1977;134:263–269.
Ziegler KM, Pitt HA, Zyromski NJ et al. Choledochoceles: are they choledochal cyst?
Ann Surg 2010, 252(4): 683–690.
0002086393.INDD 331 7/14/2014 8:04:08 PM

27 Biliary Malignancies
Michael G. House
Cancers arising from the gallbladder and biliary tree are uncommon and
often present with advanced, incurable stages. The lack of effective systemic
and targeted therapies against these cancers has also added to the medical
nihilism associated with biliary malignancy in general. Most histopatho-
logic subtypes of adenocarcinoma along the biliary tract (e.g., gallbladder,
extrahepatic [distal, perihilar], intrahepatic) carry infiltrative biologic
behavior and long-term survival, even for individuals who undergo com-
plete, potentially curative operative resection is exceptionally uncommon.
In the United States, approximately 10,000 new cases of biliary tract
cancer are diagnosed each year, with the highest incidence among Native
American and Hispanic women. Adenocarcinoma of the gallbladder stands
as the most common malignancy of the biliary tree with an overall inci-
dence of 1.5 per 100,000. Similar to all biliary tract cancers, gallbladder
cancer incidence varies dramatically according to racial and geographic
groups. Worldwide, high incidences of gallbladder cancer are reported in
women living in India, Pakistan, Ecuador, and Chile.
Over the past three decades, the incidence of intrahepatic cholangio-
carcinoma has increased worldwide while the incidence of extrahepatic
bile duct cancers, excluding gallbladder cancer, has shown a downward
trend. Several explanations for these observations have been hypothesized
and include more accurate classification systems, improvements in diag-
nostic imaging, changes in etiologic factors, and greater health care access
for low socioeconomic groups.
ETIOLOGY
Although most patients with gallstones do not develop cancer, the most
common factor associated with gallbladder cancer is gallstones leading
to chronic cholecystitis. Past studies have reported that large gallstones
(i.e., 3 cm or larger) may carry a 10-fold relative risk of gallbladder cancer.
Calcification of the gallbladder wall (i.e., porcelain gallbladder) reflecting
chronic inflammation is also associated with malignancy. Historically, this
risk was reported to be as high as 50%, but more accurately is estimated at
less than 10%. Other chronic inflammatory diseases of the biliary tract, for
example, anomalous pancreatobiliary junction, cholecystoenteric fistula,
and typhoid bacillus, are associated with gallbladder cancer and biliary
malignancy in general. While chronic inflammation is one predisposing
factor for biliary malignancy, exposure to carcinogens (e.g., thorium, radon,
dioxin, nitrosamines, vinyl chloride, isoniazid, and asbestos) and bile com-
position rich in free radicals may further potentiate malignant transforma-
tion of the biliary epithelium.
Similar to other gastrointestinal cancers, gallbladder cancer seem-
ingly follows an adenoma to carcinoma sequence that reflects specific
genetic mutation events. However, polypoid-type gallbladder cancers are
332
002113110.INDD 332 7/14/2014 8:08:13 PM

Chapter 27 / Biliary Malignancies 333
extremely rare, and most patients with true multifocal polyps are not at
increased risk for cancer development. Cholecystectomy is recommended
for patients with true adenomatous polyps as the cancer incidence of mal-
ingancy among cholecystectomy specimens harboring polyps greater than
1 cm has been reported to be as high as 10%.
The most common risk factor associated with bile duct cancer in the
United States is primary sclerosing cholangitis (PSC). The natural history of
PSC varies, and the true incidence of cholangiocarcinoma within this popu-
lation is difficult to calculate. An estimated 8% of PSC patients, diagnosed
for at least 5 years, eventually develop cholangiocarcinoma; however, nearly
40% of liver explants at the time of transplantation contain occult bile duct
cancer. The risk of developing cholangiocarcinoma in patients with ulcer-
ative colitis-associated PSC is estimated at 30% at 20 years.
Congenital choledochal cysts are associated with an increased risk of
cholangiocarcinoma especially in individuals with an anomalous pancrea-
tobiliary junction as the risk of bile duct cancer may approach 15% in this
population and appears to be time dependent. Chronic inflammation of
the bile ducts as a result of pancreatic juice reflux is felt to be the etiologic
factor, and similar risk has been observed in patients who undergo biliary
sphincteroplasty or sphincterotomy. Outside of Caroli disease, biliary cyst-
adenocarcinoma arising from a chronic, untreated cystadenoma is rare.
The risk of developing bile duct cancer, especially intrahepatic, is
also increased in patients with underlying cirrhosis secondary to chronic
hepatitis B or C infection. This finding might possibly explain the increased
incidence of intrahepatic cholangiocarcinoma in Western populations over
the past two decades. Recurrent pyogenic cholangiohepatitis and hepatoli-
thiasis resulting from chronic portal phlebitis have been established as eti-
ologies for cholangiocarcinoma, predominantly intrahepatic, in Southeast
Asia. Biliary parasites, namely Clonorchis sinensis and Opisthorchis viverrini,
which are endemic in various regions of Asia, increase the risk for both
extra- and intrahepatic cholangiocarcinoma.
DIAGNOSIS
Three clinical situations for the diagnosis of gallbladder cancer exist: final
pathology of a routine cholecystectomy specimen reveals an incidental
cancer, cancer is discovered incidentally at the time of cholecystectomy, or
gallbladder cancer is detected preoperatively. This latter situation is often
associated with an advanced stage of disease and should be suspected in
patients with right upper quadrant pain, weight loss, anorexia, and jaun-
dice. Except for advanced stages of gallbladder cancer, laboratory testing
of serum is not helpful for diagnosis. Serum CA 19-9 levels are sensitive but
not specific for gallbladder cancer. Since gallbladder cancer is often asymp-
tomatic in early stages, the majority of resected cases are diagnosed inci-
dentally with routine cholecystectomy.
In the modern era of readily available imaging techniques, for exam-
ple, ultrasound and computed tomography (CT), most patients diagnosed
with gallbladder cancer have preoperative imaging available. Early stages
of gallbladder cancer can be detected with dedicated ultrasound that may
help to delineate echogenic mucosa associated with carcinoma compared
to benign disease and the presence of modest hepatic parenchymal inva-
sion. Cross-sectional imaging, with either CT or MRI, is crucial during
the diagnostic and staging evaluation of suspected gallbladder cancer
preoperatively or for complete staging when the diagnosis is established
intra- or postoperatively. Both modalities provide excellent resolution for
determining the local extent of disease (hepatic, vascular, and adjacent
0002113110.INDD 333 7/14/2014 8:08:13 PM

organ involvement) and the presence of local or regional lymphadenopathy.

Distant sites of metastasis are best detected with CT imaging, which may
be enhanced with positron emission tomography (PET) using fluorodeoxy-
glucose (FDG).
The diagnosis of intrahepatic cholangiocarcinoma is typically estab-
lished when a hypovascular tumor is detected on CT or MRI. Histologic con-
firmation is not necessary unless resection is not an option for treatment
or the tumor carries atypical radiographic features. Hepatic metastasis
from a primary source from the gastrointestinal tract, breast, lung, kidney,
etc. can be excluded with upper and lower endoscopy and/or whole-body
PET-CT functional imaging. Complete staging to evaluate for intra- and
extrahepatic sites of metastasis (e.g., regional lymph nodes, peritoneum)
and major vascular invasion can be accomplished with contrast-enhanced
cross-sectional imaging (CT or MRI).
The diagnosis of hilar and distal cholangiocarcinoma usually follows
targeted studies for the evaluation of obstructive jaundice or elevated liver
enzymes. Most patients are referred after having had some initial studies
done elsewhere, usually CT and some form of direct cholangiography, for
example, endoscopic retrograde cholangiopancreatography (ERCP). Just as
for gallbladder cancer, histologic confirmation of malignancy is not manda-
tory before planning operative treatment. In the absence of previous biliary
tract surgery, the finding of a focal stenotic lesion along the extrahepatic
biliary ducts combined with the appropriate clinical presentation (i.e.,
jaundice) is adequate for a presumptive diagnosis of hilar or distal cholan-
giocarcinoma depending on the level of stricture formation.
Distal bile duct tumors frequently are mistaken for adenocarcinoma
of the pancreatic head, the most common periampullary malignancy. ERCP
and (magnetic resonance cholangiopancreatography) MRCP are typically
used to evaluate periampullary tumors when the diagnosis is suspected. As is
true for hilar lesions, MRCP can provide images of the distal bile duct previ-
ously obtainable only with ERCP or percutaneous cholangiography. A dilated
extrahepatic bile duct terminating abruptly at its distal aspect without a con-
comitantly dilated pancreatic duct suggests a distal bile duct cancer.
Although most patients with hilar strictures and jaundice have
cholangiocarcinoma, alternative diagnoses are possible in up to 15% of
patients. The most common of these are gallbladder carcinoma, Mirizzi
syndrome, and benign focal strictures (e.g., lymphoplasmacytic sclerosing
pancreatitis/cholangitis, granulomatous disease, and PSC). The finding of a
smooth, tapered stricture on cholangiography suggests a benign stricture.
Diagnostic assessments based on the cholangiographic appearance of the
stricture are unreliable, and hilar cholangiocarcinoma must remain the
leading diagnosis until disproved definitively. Relying on the results of per-
cutaneous or endoscopic ultrasound needle biopsy or biliary brush cytol-
ogy is not recommended because these results can be misleading. Many
tumors express carcinoembryonic antigen (CEA) and carbohydrate antigen
19-9 (CA 19-9); however, serum levels of these markers, although elevated in
some patients, often have little diagnostic value. Similarly, elevations of CA
19-9 or CEA levels in intraductal bile are not reliable or accurate.
The diagnosis of malignancy in patients with bile duct strictures
can be challenging and depends on accurate interpretation of ERCP as
well as cross-sectional imaging findings. ERCP, in particular, is a useful
tool for assessing biliary tract strictures as it permits visualization of the
biliary tree with the opportunity for therapeutic interventions (i.e., bili-
ary drainage) and collection of specimens for cyto- and histopathologic
evaluation. Biliary brushing cytology, although highly specific, has suf-
fered from low to moderate sensitivity (15% to 68%). As a result, adjunctive
002113110.INDD 334 7/14/2014 8:08:14 PM

iagnostic techniques including mutation analysis, DNA ploidy analysis,

d
and fluorescence in situ hybridization (FISH) have been studied clinically
and experimentally to improve tumor detection sensitivity. Two types of
chromosomal abnormalities are frequently identified with a clinical FISH
probe set, namely polysomy and trisomy 7. Polysomy has been defined as a
gain of two or more of the four probes in ≥5 cells. FISH is now considered
more sensitive than cytology for detecting biliary tract cancer.
In patients with a stricture of the extrahepatic bile duct and a clini-
cal presentation consistent with cholangiocarcinoma or gallbladder car-
cinoma, histologic confirmation of malignancy is generally unnecessary,
unless nonoperative therapy is planned. Endoscopic brushings of the bile
duct have an unacceptably low sensitivity, making a negative result virtually
useless. Excessive reliance on the results of endoscopic ultrasound-guided
fine needle aspiration or endoscopic brush biopsies serves only to delay
therapy.
The diagnosis of cancer in the setting of PSC can be difficult as a result
of chronic liver disease, diffuse biliary strictures, and inflammatory changes
of the biliary epithelium. New-onset jaundice, elevations of serum CA 19-9
level, clinical weight loss, dominant biliary stricture formation with or with-
out cytologic or aneuploidic changes of dysplasia, or a detectable mass on
cross-sectional imaging may signify carcinoma.
MANAGEMENT
Preoperative Staging
Radiographic studies are pivotal in selecting patients for resection of
bile duct cancers. In addition to providing accurate preoperative staging
information and anatomic delineation of local disease extent, cross-
sectional imaging allows for careful operative planning. CT is an impor-
tant study for evaluating patients with biliary obstruction. A high-quality
CT scan can provide valuable information regarding the level of obstruc-
tion, vascular involvement, and liver atrophy. Advances in CT technology
permits the acquisition of three distinct circulatory phases, consisting of
the arterial, early, and late venous phases, in the hilar and pancreatobili-
ary regions.
Patients with hilar cholangiocarcinoma are difficult to stage preop-
eratively. The modified Bismuth-Corlette classification stratifies patients
based on the extent of biliary duct involvement by tumor alone (Fig. 27.1).
Although useful for operative planning, it does not predict resectability or
survival. The American Joint Commission for Cancer (AJCC) TNM stage
system (Table 27.1) is based largely on pathologic criteria and has little
relevance for preoperative staging. A preoperative staging system should
predict operative resectability and survival accurately. Preoperative
assessments should include (1) the extent of tumor within the biliary tree,
(2) vascular involvement, and (3) lobar atrophy (Table 27.2). This clinical
staging scheme underscores the importance of considering portal vein
involvement and liver atrophy in relation to the extent of ductal cancer
spread. Ipsilateral involvement of vessels and bile ducts is usually ame-
nable to resection, whereas contralateral involvement is not. Unilateral
involvement of the hepatic artery is compatible with resection. Bilateral
involvement of the hepatic artery usually is associated with nonresect-
ability when the tumor is located mainly on the right. Hepatic arte-
rial reconstruction in order to accomplish a complete resection should
be reserved for highly select patients with potentially curable disease.
Similarly, main portal vein invasion, while not an absolute contraindica-
tion for resection of hilar cholangiocarcinoma, will preclude a complete
0002113110.INDD 335 7/14/2014 8:08:14 PM

Type I Type II
A Type III B
FIGURE 27.1 Bismuth-Corlette classification scheme of malignant biliary strictures.
resection with long-term benefit for the majority of patients. Unresectable

hilar cholangiocarcinoma has been considered for orthotopic liver trans-
plantation in highly selected patients with protocolized conditions and
treatments.
The role of fluorodeoxyglucose positron emission tomography (FDG-
PET) in the preoperative staging of patients with carcinoma of the gallblad-
der or biliary tract is unclear. The putative benefits of FDG-PET to identify
occult metastatic disease or advanced regional lymphadenopathy have been
difficult to realize in clinical practice that utilizes timely, high-resolution
cross-sectional imaging routinely.
Preoperative Planning
High-resolution MRI cholangiopancreatography and direct cholangi-
ography (e.g., ERCP or percutaneous cholangiography) provide impor-
tant anatomic information, including the location of the tumor and the
biliary extent of disease, both of which are crucial in operative planning.
High-quality cholangiograms are essential for accurately diagnosing the
extent and level of bile duct involvement. High-performance MRCP has
almost replaced endoscopic and percutaneous cholangiography as the
initial preoperative assessment of hilar cholangiocarcinoma. MRCP not
only identifies the tumor and the level of biliary obstruction but also may
reveal obstructed and isolated ducts not appreciated at endoscopic or
002113110.INDD 336 7/14/2014 8:08:14 PM

TABLE
AJCC Staging System, Seventh Edition, for Hilar Bile Duct
27.1 Cancer
Primary Tumor (pT)

pTX: Cannot be assessed
pT0: No evidence of primary tumor
pTis: Carcinoma in situ
pT1: Tumor confined to the bile duct, with
extension up to the muscle layer or fibrous
tissue
pT2a: Tumor invades beyond the wall of the bile
duct to surrounding adipose tissue
pT2b: Tumor invades adjacent hepatic parenchyma
pT3: Tumor invades unilateral branches of the
portal vein or hepatic artery
pT4: Tumor invades main portal vein or its
branches bilaterally; or the common
hepatic artery; or the second-order biliary
radicals bilaterally; or unilateral second-
order biliary radicals with contralateral
portal vein or hepatic artery involvement
Regional Lymph Nodes (pN)
pNX Cannot be assessed
pN0 No regional lymph node metastasis
pN1 Regional lymph node metastasis (including
nodes along the cystic duct, common bile
duct, hepatic artery, and portal vein)
pN2 Metastasis to periaortic, pericaval, superior
mesentery artery, and/or celiac artery
lymph nodes
Distant Metastasis (pM)
pMx Cannot be assessed
pM1 Distant metastases present
Stage Groupings
Stage 0 Tis N0 M0
Stage I T1 N0 M0
Stage II T2a or T2b N0 M0
Stage IIIA T3 N0 M0
Stage IIIB T1, T2, or T3 N1 M0
Stage IVA T4 N0 or N1 M0
Stage IVB Any T N2 M0 or M1
Any T Any N M1
From Cholangiocarcinoma. In: Edge SB, Byrd DR, Compton CC (eds.). AJCC Cancer Staging Manual,
7th ed. New York, NY: Springer, 2010, with permission.
percutaneous study (Fig. 27.2). MRCP also provides information regarding

the patency of hilar vascular structures, the presence of nodal or distant
metastases, and the presence of lobar atrophy. Ultrasound, albeit opera-
tor dependent, is another useful noninvasive study during the preoperative
planning process. When utilized by experienced centers, it often delineates
tumor extent accurately. Ultrasound not only shows the level of biliary duc-
tal obstruction but also can provide information regarding tumor exten-
sion within the bile duct and in the periductal tissues. Duplex ultrasound
0002113110.INDD 337 7/14/2014 8:08:14 PM

TABLE
27.2
Stage
Preoperative Staging System for Hilar Cholangiocarcinoma
Criteria
T1 Tumor involving biliary confluence ± unilateral extension to
2 biliary radicles
T2 Tumor involving biliary confluence ± unilateral extension to
2 biliary radicles
and ipsilateral portal vein involvement ± ipsilateral hepatic
lobar atrophy
T3 Tumor involving biliary confluence + bilateral extension to
2 biliary radicles
Or unilateral extension to 2 biliary radicles with contralateral
portal vein involvement
Or unilateral extension to 2 biliary radicles with contralateral
hepatic lobar atrophy
Or main or bilateral portal venous involvement
From Jarnagin WR, Fong Y, DeMatteo RP, et al. Staging, resectability, and outcome in 225 patients
with hilar cholangiocarcinoma. Ann Surg 2001;234:507–519, with permission.
is firmly established as a highly accurate predictor of vascular involvement

and technical resectability.
ERCP, when performed skillfully, provides useful ductal detail and
can effectively drain functional liver segments prior to operative plan-
ning. Percutaneous cholangiography, while invasive, outlines the extent
of intrahepatic bile duct involvement and has been the preferred study in
the past for patients with hilar cholangiocarcinoma. Compared to endo-
biliary drainage, PTC may provide more reliable biliary drainage of the
functional liver remnant when major hepatectomy is planned along with
biliary resection. The benefits of preoperative biliary drainage have been
difficult to prove, but most centers currently advocate for such interven-
tion prior to resection of the extrahepatic bile ducts along with major
hepatectomy.
The absolute indications for preoperative biliary drainage for jaun-
diced patients prior to pancreatoduodenectomy are not clear. Randomized
controlled trials fail to demonstrate an improvement in postoperative
outcomes for patients with periampullary cancers; however, most experts
continue to advocate for preoperative endobiliary drainage for patients
with deep jaundice, that is, serum total bilirubin greater than 15 mg/dL.
Preoperative ERC with endobiliary drainage is the preferred approach in
this situation.
For patients with hilar and intrahepatic cholangiocarcinoma who
will require major hepatectomy, three-dimensional CT offers the capacity
to generate volumetric models of the total liver, planned hepatic resection,
and functional remnant liver volume. Since patients subjected to resec-
tion of more than 70% of functional liver parenchyma have substantially
a higher risk of postoperative liver failure regardless of existing cholestatic
liver disease, routine preoperative volumetry may help to identify patients
who may be at considerable risk for postoperative hepatic failure and select
candidates for preoperative portal vein embolization. The purpose of portal
vein embolization is to increase the functional liver mass by inducing com-
pensatory hypertrophy in the future liver remnant with the ultimate goal
to minimize postoperative liver dysfunction that contributes to mortality
substantially.
002113110.INDD 338 7/14/2014 8:08:15 PM

FIGURE 27.2 Coronal (A) and axial (B) MRCP images of a patient with hilar cholangio-
carcinoma. The tumor involves the right and left hepatic ducts. The bile ducts in this study
appear white. Extreme atrophy of the left hemiliver is apparent.
In right portal vein embolization, the volume of the left lobe increases
by 130 cm3 on average within 2 weeks after embolization, and the esti-
mated resection volume decreases by an average of approximately 10%.
In the absence of randomized controlled trials, the indications for portal
vein embolization remain somewhat arbitrary, but many centers utilize
this technique preoperatively for patients undergoing extended right
0002113110.INDD 339 7/14/2014 8:08:15 PM

epatectomy in the presence of underlying liver disease. In recent years,

h
some centers have experimented with associating liver partition and portal
vein ligation for staged hepatectomy (ALPPS) as a perioperative technique
to induce rapid hypertrophy of the future liver remnant and avoid postop-
erative liver dysfunction.
OPERATIVE INDICATIONS AND THERAPY

Staging laparoscopy for biliary tract malignancy should be considered for
all patients with suspicious clinicopathologic findings for metastatic dis-
ease (e.g., remarkable elevations of serum CA 19-9 level, extreme weight
loss, rapidly declining performance status, indeterminant nodules involving
the liver or peritoneum). With an overall yield of 30%, staging laparoscopy
should be utilized for all patients with gallbladder cancer and intrahepatic
cholangiocarcinoma involving the liver capsule. Since the detection rate for
occult metastases is stage dependent, most surgeons apply staging laparos-
copy selectively for patients with extrahepatic bile duct cancer.
For gallbladder cancer in general, operative resection is indicated for
patients with localized disease, that is, tumors limited to the gallbladder,
liver bed, and regional lymph nodes. Peritoneal dissemination, extensive
liver or bile duct (especially left hepatic duct) involvement, nodal metas-
tases beyond the hepatoduodenal ligament (i.e., N2 disease, Fig. 27.3), left
hepatic artery or portal vein encasement, or pancreatoduodenal infiltration
FIGURE 27.3 Regional lymph nodes associated with gallbladder cancer. The subhilar N1
level lymph nodes within the hepatoduodenal ligament are encircled. Lymph nodes out-
side of the circle (e.g., celiac, aortocaval, and mesenteric nodes) are considered N2 level.
(Figure borrowed from Lillemoe K, Jarnagin WR (eds.). Hepatobiliary and pancreatic surgery.
002113110.INDD 340 7/14/2014 8:08:17 PM

should serve as contraindications for reasonable attempts at curative resec-

tion. Intraoperative ultrasound is a useful tool to further delineate spatial
relationships between the tumor, liver, and surrounding hilar structures.
Tumors limited to the mucosa (T1a) can be treated sufficiently with
cholecystectomy alone. Patients with deeper invasion require en bloc par-
tial hepatectomy (wedge resection of adjacent liver for T1b tumors; seg-
ment IVb + V resection for T2-3 tumors) with regional subhilar periportal
lymphadenectomy. Resection of the extrahepatic bile ducts followed by
reconstruction with hepaticojejunostomy should be reserved for tumors
that extend into the hepatic or common bile duct. Tumor invasion of the
right hepatic duct and/or right hepatic artery is not a contraindication to
operative resection, which requires an extended right hepatectomy in this
circumstance.
Appreciating the extent (size, multifocality, major vascular involve-
ment) of intrahepatic bile duct cancer on preoperative imaging and intra-
operative ultrasound is crucial for planning hepatic resections for this
disease. These tumors frequently arise from the segment IV ducts and
involve the central liver inflow anatomy. Three phenotypes of intrahepatic
cholangiocarcinoma have been described: periductal infiltrating, mass-
forming, and intraductal/papillary. Usually large, these tumors often
require a major hepatectomy that warrants serious preoperative con-
sideration for the size and status of the future liver remnant (described
above under preoperative planning). Proper incision and exposure are
essential elements of a safe hepatectomy. Hemostatic hepatic parenchy-
mal transection has been aided by several energy devices, including the
ultrasonic aspirator, saline-linked cautery, bipolar electrocoagulation,
and ultrasonic scalpel. A solitary tumor located peripherally, especially
within an anterior segment, is often amenable to a safe laparoscopic tech-
nique that focuses on segmental anatomy for optimal tumor clearance.
The spatial relationship of the tumor to the sectoral bile ducts and hepatic
ducts needs to considered in all cases. A periductal infiltrating phenotype
of a tumor with involvement of the left, right, or common hepatic ducts
will require extrahepatic biliary resection often with the need for biliary
reconstruction.
The Bismuth-Corlette classification scheme provides a rudimentary
guide for planning operative resection for hilar cholangiocarcinoma. Type I
tumors without involvement of the right hepatic artery can be treated with
resection of the entire extrahepatic biliary tree (without hepatectomy), sub-
hilar lymphadenectomy, and bilateral cholangiojejunostomy. Since biliary
branches of the hepatic caudate are involved in most cases of type II-III hilar
cholangiocarcinoma, caudate resection en bloc with a major hepatectomy
and extrahepatic bile duct resection is indicated for these tumors. Type IIIa
tumors, which may encase the right hepatic artery, are approached with
an extended right hepatectomy or formal right trisectionectomy with cau-
date resection. In the absence of right hepatic artery involvement, type IIIb
tumors are resected with extrahepatic bile duct excision, left hepatectomy,
and caudate resection. The final bile duct margin status, both proximally
and distally, should be assessed intraoperatively with frozen section histopa-
thology. Sub- and perihilar lymphadenectomy should be considered routine
for all operations. Kocherization of the duodenum and head of the pancreas
during early exploration permits examination of the retropancreatic and
aortocaval spaces to exclude advanced levels of lymph node metastases and
limited opportunity for long-term benefit following operative resection.
Pancreatoduodenectomy for periampullary carcinoma has been
described extensively throughout this text. Understanding the longitudi-
nally spreading nature of cholangiocarcinoma, most surgeons recommend
0002113110.INDD 341 7/14/2014 8:08:17 PM

a high transection of the common hepatic duct in patients with distal bile
duct cancer. The proximal hepatic duct margin should be analyzed intraop-
eratively to ensure tumor clearance.
PREVENTION AND TREATMENT OF POSTOPERATIVE

COMPLICATIONS
Postoperative fluid collections after hepatectomy are usually related to
either bleeding or bile leaks from the transected liver surface. Intraoperative
transcystic leak tests after hepatectomy can be performed, but have not
been shown to affect the incidence of postoperative bile leaks. Leaks from
small biliary radicals rarely require intervention; however, 8% of all hepa-
tectomies result in bile leaks from segmental ducts requiring percutaneous
drainage. Percutaneous drainage alone will treat the majority of perihepatic
fluid collections. Persistent bile leaks after percutaneous drainage should
raise suspicion for a stricture of the native distal bile ducts or biliary–
enteric anastomosis. In these situations, endoscopic or percutaneous bili-
ary stenting may be necessary for resolution.
The reported association between deep organ space infection after
hepatectomy and biliary–enteric anastomosis encourages judicious opera-
tive drain placement. Decompression of the biliary–enteric anastomosis
with transhepatic stenting has been utilized by some surgeons to decrease
the incidence of infectious complications after hepatectomy.
The risk of postoperative hepatic failure after major hepatectomy is
related to the volume and quality of the functional liver remnant. Factors
that predispose patients to postoperative hepatic failure include age,
gender, diabetes mellitus, fatty liver disease, operative blood loss, small
liver remnant (<20% normal liver volume, <40% cirrhotic liver volume),
cholestasis, and postoperative infection. Preoperative factors can be
addressed with techniques to improve the functional liver remnant, for
example, biliary drainage and portal vein embolization. Postoperative
liver failure must be addressed promptly in the postoperative period with
drainage of perihepatic fluid collections and appropriate treatment of
infection.
OUTCOMES
Stage-dependent survival after resection for gallbladder adenocarcinoma
is summarized in Figure 27.4. Median survival for stage IV gallbladder can-
cer is less than 6 months; thus, operative intervention, even with palliative
intent, for this stage of disease is never advocated. Simple cholecystectomy
for pathologically staged T1a cancer results in cure in most cases, whereas
more advanced stages of disease requiring extensive operations for tumor
clearance lend themselves to long-term survival infrequently. The role of
adjuvant therapy for gallbladder adenocarcinoma remains unclear. Past
studies of 5-fluorouracil or gemcitabine with or without external beam
radiotherapy failed to show a benefit in long-term survival; however, the
addition of cisplatin to gemcitabine has improved treatment response rates
for unresectable biliary tract cancers considerably. Future collaborative tri-
als are necessary to study the benefits of adjuvant platinum-based chemo-
therapy for bile duct cancer.
Aggressive operations for hilar cholangiocarcinoma have been
associated with major postoperative complications and mortality. But,
improvements in operative technique along with better patient selec-
tion and avoidance of postoperative hepatic failure have decreased
postoperative morbidity and mortality substantially over the past two
decades. Long-term survival is realized only for patients who undergo
002113110.INDD 342 7/14/2014 8:08:17 PM

100
Stage 0
80
60 I
40
II
20 IIIA/IIIB
0 IVA/IVB
Dx 1Y 2Y 3Y 4Y 5Y
FIGURE 27.4 Overall 5-year survival for resected gallbladder adenocarcinoma, accord-
ing to AJCC seventh edition cancer stage. (Figure borrowed from Lillemoe K, Jarnagin
WR (eds.). Hepatobiliary and pancreatic surgery. Philadelphia, PA: Lippincott Williams &
Wilkins, 2013.)
an R0 resection and do not carry lymph node metastases. Orthotopic

liver transplantation for hilar cholangiocarcinoma can achieve excellent
long-term outcomes for highly selected patients (i.e., no lymph node or
distant metastases) who demonstrate no evidence of disease progression
with neoadjuvant chemoradiotherapy. Liver transplant for cholangiocar-
cinoma is only performed in select centers with extremely close clinical
selection criteria.
Patients with unresectable, locally advanced hilar tumors but without
evidence of widespread disease may be candidates for palliative radiation
therapy. A combination of stereotactic external beam radiation (5,000 to
6,000 cGy) with or without intraductal iridium-192 (2,000 cGy) delivered
percutaneously or endoscopically can be applied. Intraluminal (either
endoscopic or transhepatic) photodynamic therapy has been utilized for
unresectable hilar cholangiocarcinoma. This two-step procedure involves
systemic administration of a photosensitizer followed by direct illumina-
tion via cholangioscopy, which activates the compound causing local tumor
cell death. Two randomized studies in patients with unresectable cholan-
giocarcinoma suggested improved survival with biliary stenting combined
with photodynamic therapy compared with biliary stenting alone, but the
improved survival observed with photodynamic therapy is likely related to
better biliary decompression and avoidance of early segmental duct isola-
tion and subsequent cholangitis, rather than any significant reduction in
tumor burden.
Several pathologic features of intrahepatic cholangiocarcinoma
have adverse effects on prognosis even after complete resection. Lymph
node metastases, multifocality, and major vascular invasion all have nega-
tive impact on survival. Liver transplantation has been utilized for highly
selected patients with technically unresectable tumors confined to the liver.
Transarterial chemoembolization and radioembolization can be offered to
patients with unresectable disease in an attempt to slow disease progres-
sion and possibly improve survival.
0002113110.INDD 343 7/14/2014 8:08:18 PM

CONCLUSION
Enhanced radiologic imaging and greater understanding of malignancies of
the biliary tract have permitted improved operative outcomes over the past
few decades. Progress in these two areas has allowed better patient selec-
tion, preoperative preparation of the future liver remnant, and techniques
to treat bile duct cancer from a surgical standpoint. Even in the modern
era, highly effective systemic agents for cholangiocarcinoma are lacking,
which leaves a large therapeutic gap for patients with advanced stages of
disease. One of the main goals of treating patients with bile duct cancer is
palliation from biliary obstruction, which can be accomplished endoscopi-
cally, percutaneously, or operatively. With recent improvements in chemo-
therapy, endoluminal therapy, and brachyradiotherapy, a multidisciplinary
approach toward treating bile duct cancer should be advocated in order to
achieve optimal individualized treatments for patients.
Suggested Readings
D’Angelica M, Fong Y, Weber S, et al. The role of staging laparoscopy in hepatobiliary
malignancy: prospective analysis of 401 cases. Ann Surg Oncol 2003;10:183–189.
Endo I, House MG, Klimstra D, et al. Clinical significance of intraoperative bile
duct margin assessment for hilar cholangiocarcinoma. Ann Surg Oncol 2008,
15(8):2104–2112.
Farges O, Belghiti J, Kianmanesh R, et al. Portal vein embolization before right hepa-
tectomy: prospective clinical trial. Ann Surg 2003; 237:208–217.
Jarnagin WR, Fong Y, DeMatteo RP, et al. Staging, resectability, and outcome in 225
patients with hilar cholangiocarcinoma. Ann Surg 2001;234:507–519.
Kawasaki S, Imamura H, Kobayashi A. Results of surgical resection for patients with
hilar bile duct cancer: application of extended hepatectomy after biliary drainage
and hemihepatic portal vein embolization. Ann Surg 2003;238:84–92.
Kennedy TJ, Yopp A, Qin Y, et al. Role of preoperative biliary drainage of liver rem-
nant prior to extended liver resection for hilar cholangiocarcinoma. HPB
2009;11(5):445–451.
Loehrer AP, House MG, Nakeeb A, et al. Cholangiocarcinoma: Are North American
outcomes optimal? J Am Coll Surg 2013;216(2):192–200.
Nagino M, Kamiya J, Arai T, et al. Two hundred forty consecutive portal vein embo-
lizations before extended hepatectomy for biliary cancer: surgical outcome and
long-term follow-up. Ann Surg 2006;243:364–372.
Quyn AJ, Ziyaie D, Polignano FM, Tait IS. Photodynamic therapy is associated with an
improvement in survival for patients with irresectable hilar cholangiocarcinoma.
HPB 2009;11(7):570–577.
Rea DJ, Heimbach JK, Rosen CB, et al. Liver transplantation with neoadjuvant chemo-
radiation is more effective than resection for hilar cholangiocarcinoma. Ann Surg
2005; 242:451–458.
Schnitzbauer AA, Lang SA, Goessmann H, et al. Right portal vein ligation combined
with in situ splitting induces rapid left lateral liver lobe hypertrophy enabling
2-staged extended right hepatic resection in small-for-size settings. Ann Surg
2012;255(3):405–414.
Valle J, Wasan H, Palmer DH, et al. Cisplatin plus gemcitabine versus gemcitabine for
biliary tract cancer. NEJM 2010;362(14):1273–1281.
Van der Gaag, NA, Rauws EA, van Eijck CH, et al. Preoperative biliary drainage for
cancer of the head of the pancreas. NEJM 2010;362(2):129–137.
002113110.INDD 344 7/14/2014 8:08:18 PM

28 Minimally Invasive Biliary
Surgery
Eugene P. Ceppa
Minimally invasive biliary surgery is indicated in only a few isolated

complex biliary disease processes when technical expertise is available.
Historically and currently, the vast majority of biliary surgery is performed
through an open approach. However, the benefits of laparoscopic surgery
have been established in numerous other areas of general surgery includ-
ing cholecystectomy, gastric bypass, and colectomy. Currently, a paucity of
data exists regarding the role of minimally invasive biliary surgery. Through
the improvement of surgical skill (experience and fellowship training) and
technology (optics, robotics, surgical staplers, and tissue-sealing devices),
surgeons have begun to challenge the standard of care in biliary surgery.
In order to avoid overlap, this chapter focuses primarily on laparoscopic
Roux-en-Y hepaticojejunostomy as the preferred biliary–enteric recon-
struction for the treatment of biliary disease.
Preoperative considerations focus primarily on patient variables that affect
minimally invasive surgery and preparation of the patient for surgery.
Patient variables to consider include comorbid conditions, past surgical
history (especially abdominal surgery), body habitus, and hepatoduode-
nal ligament anatomy. Comorbid conditions play a role in laparoscopic
approach primarily as a function of length of anesthetic time and ability to
clear carbon dioxide from the pneumoperitoneum. Patients with advanced
cardiopulmonary disease are particularly fragile if the length of anesthetic
time is excessive. This may negate the typical benefits of the laparoscopic
approach. As a surgeon gains experience, operative times typically decrease
dramatically, yet the learning curve in complex biliary surgery is unknown.
Airway pathology (obstructive sleep apnea) and pulmonary disease (chronic
obstructive pulmonary disease) can be exacerbated as a result of prolonged
anesthetic and subsequent insufflation with carbon dioxide. Nearly 100%
of patients who possess a body mass index (BMI) of greater than 45 have
obstructive sleep apnea (with or without documented sleep studies), and
patients with home oxygen requirements or room air arterial CO2 concen-
tration of greater than 44 mmHg should be identified as high risk prior to
entering the operative theater. A history of portal hypertension or cirrhosis
can be a major deterrent regardless of approach.
Past surgical history is a major component for surgeon decision mak-
ing in terms of peritoneal access, port placement, the ability to perform the
jejunojejunostomy, and whether the Roux limb will reach the porta hepatis
for bile duct reconstruction. History taking and examination are critical
to determine previous abdominal operations and location of concurrent
cicatrices. Ventral hernia repair with mesh should also be specifically
determined as peritoneal access and contamination of synthetic mesh with
bile can create new infectious problems. More extensive open abdominal
345
0002086389.INDD 345 7/14/2014 8:12:28 PM

perations tend to have more robust adhesions around the small bowel,
o
which can either prevent formation of an adequate Roux limb or increase
operative times for the required enterolysis.
Body habitus is not a contraindication to laparoscopy but deters some
surgeons for multiple reasons. Larger patients (BMI > 35) require longer
instruments; thus, reach to the porta hepatis can be difficult. Abdominal
wall tension on laparoscopic instruments can place undue stress to long,
thin instruments and can either limit the ability to use some necessary
instruments or damage them beyond functional use. Mobilization of the
hepatic flexure and a Kocher maneuver are made more difficult when the
transverse mesocolon and omentum contain more fat and organs are larger
in size overall; once mobilized, the hepatic flexure and the retroperitoneum
lateral to the duodenum can be more difficult to retract for exposure pur-
poses when fatty or large. Finally, super morbidly obese (BMI > 50) patients
typically have engorged livers as a result of fatty liver disease and are prone
to lacerations (hemorrhage within the operative field significantly impairs
image brightness) and fractures leading to further surgery to correct the
problem. These patients should be considered for a preoperative liquid diet
with protein supplementation (7 days prior) to diminish the water weight
of the liver resulting in a temporary reduction in liver size to optimize the
intraoperative exposure and minimize surrounding tissue injury.
Mastering the patient’s preoperative diagnostic imaging can make
intraoperative decision making clearer. Cholangiography provides an intra-
luminal roadmap of the biliary tree. Identification of cystic duct or cystic
duct clips can be performed safely and provide intraoperative details about
the location of the bile duct. Also, knowledge of bile duct anomalies helps
avoid injury when they are expected prior to surgery. Cross-sectional imag-
ing with arterial and venous phases assists in determining the location of
the hepatic arterial branches (specifically if the right hepatic artery tra-
verses anterior to or passes posterior to the common hepatic duct to avoid
inadvertent injury or ligation). Replaced right or completely replaced com-
mon hepatic artery branches should be identified with imaging to avoid
injury during mobilization of the bile duct. Portal vein, superior mesenteric
vein, and splenic vein thrombosis should be sought out prior to surgery as
collateral veins can make the hepatoduodenal ligament dissection treach-
erous. Finally, a large, pathologically dilated bile duct can be easily seen on
diagnostic imaging, and some surgeons will only attempt a laparoscopic
hepaticojejunostomy on dilated bile ducts as a tenet of patient selection.
Patients with complex biliary disease commonly have other medical
issues that require treatment prior to surgery. Preoperative biliary obstruc-
tion exposes the patient to the effects of jaundice, bactibilia, and external
manipulation of the biliary tree (endoscopic or transhepatic). When infection
is present in the form of cholangitis or biloma, antibiotics are indicated, and
treatment of the obstruction with biliary endoprosthesis or percutaneous bil-
iary drains and/or percutaneous peritoneal drains is required for source con-
trol. Patients with infection or biliary insensible losses may be malnourished,
may be dehydrated, possess electrolyte abnormalities, and are deficient in
vitamin K altering coagulation parameters. All of these factors should be cor-
rected prior to surgery to minimize postoperative complications.
OPERATIVE INDICATIONS AND TECHNICAL TIPS

Indications for hepaticojejunostomy are listed in Table 28.1. Biliary obstruc-
tion secondary to benign and malignant strictures is the primary indica-
tion for hepaticojejunostomy. Benign complex biliary disease includes
iatrogenic biliary stricture, choledochal cysts, chronic pancreatitis, Mirizzi
002086389.INDD 346 7/14/2014 8:12:28 PM

Chapter 28 / Minimally Invasive Biliary Surgery 347
TABLE
28.1 Indications for Biliary–Enteric Reconstruction
Benign
Iatrogenic bile duct injury
Chronic pancreatitis
Choledochal cysts
Mirizzi syndrome
Traumatic bile duct injury
Malignant
Cholangiocarcinoma (mid–bile duct)
Pancreatic adenocarcinoma—palliative
Cholangiocarcinoma (distal)—palliative
Duodenal adenocarcinoma—palliative
Ampullary adenocarcinoma—palliative
Portal lymphadenopathy—palliative
syndrome, and traumatic bile duct injuries. Choledocholithiasis and mini-

mally invasive treatment thereof can be found in Chapter 25 (common bile
duct exploration). Indications for treatment of malignant biliary strictures
include extrahepatic bile duct resection and reconstruction of cholangio-
carcinoma (Bismuth II) or palliation of periampullary tumors (pancreatic
adenocarcinoma, distal cholangiocarcinoma, duodenal adenocarcinoma,
and ampullary adenocarcinoma) and portal lymphadenopathy resulting
from metastasis of secondary malignant tumors. Periampullary tumor
treatment with curative intent via a laparoscopic pancreaticoduodenec-
tomy can be found in Chapter 8 (laparoscopic pancreatic surgery).
Most studies indicate that a Roux-en-Y hepaticojejunostomy is the saf-
est and most reproducible method of biliary–enteric reconstruction possess-
ing the best long-term patency rate. The critical skill set for this procedure
is intracorporeal suturing using laparoscopic needle drivers and graspers.
The following are the critical steps toward biliary–enteric reconstruction:
1. Patient position with both arms out is suitable.
2. Port placement can vary among surgeons, but Figure 28.1 demonstrates
an accepted configuration. The ports consist of a 5 mm port in both the
right and left subcostal margins, a 12-mm right upper quadrant port
(surgeon’s primary port), and a 5- or 12-mm port in both the left upper
quadrant and umbilicus (primarily for the camera). A subxiphoid incision
can be made without a port for placement of a liver retractor.
3. To create the Roux limb, starting in the supine position is best. The liga-
ment of Treitz is identified at the base of the transverse mesocolon.
4. A window is created between the vasa recta of the small bowel, and the
small bowel is divided using a laparoscopic stapler with a 3.5-mm staple
load approximately 50 cm distal to the ligament of Treitz.
5. Use the tissue-sealing device to divide the mesentery evenly in order not
to encroach on the arterial arcade on either side of the divided bowel
(optional). This maneuver provides a slightly longer reach to the bile duct
and avoids undue tension on the anastomosis. The bowel proximal to
the staple line is the alimentary limb; an additional 50 cm of small bowel
distal to the staple line is measured and referred to as the Roux limb.
6. A side-to-side jejunojejunostomy is fashioned:
a. Two 3-0 absorbable stay sutures are placed 6 to 10 cm apart.
b. Enterotomies are created with the cautery.
0002086389.INDD 347 7/14/2014 8:12:28 PM

5 * 5
12 5 – 12 5 – 12
FIGURE 28.1 Port site placement for laparoscopic Roux-en-Y hepaticojejunostomy.

*, Camera port; X, optional liver retractor.
c. An articulating 60-mm cartridge of a 2.5-mm staple load is intro-

duced into the bowel via the enterotomies. This load is generally
more hemostatic than is the typical 3.5-mm load used commonly for
bowel cases.
d. The bowel is manipulated so that the mesentery is outside the staple
load.
e. The anastomosis between the alimentary and Roux limbs is created
with the stapler.
f. The common enterostomy is oversewed with a 3-0 absorbable run-
ning suture.
g. The jejunojejunostomy mesenteric defect is closed with a running 2-0
nonabsorbable suture to prevent future internal hernias. Avoid full-
thickness bites across the bowel mesenteric edge that was previously
sealed with an energy device, which can lead to unnecessary hemor-
rhage, ischemia, or hematoma near the anastomosis. The subsequent
complications include bleeding, perforation, and obstruction postop-
eratively. Solely suture the peritoneum superficially together.
h. Visualize the Roux limb mesenteric edge, and travel down toward the
root of the mesentery to ensure the mesentery is not twisted (this is
worth double or triple checking as peripheral visualization is limited
during laparoscopic surgery).
i. Create a defect in the transverse mesocolon with the tissue-sealing
device to the right of the middle colic artery. Preserve the left space
in case a future pancreatic Roux limb is needed.
j. Mark the Roux limb staple line with a stitch or a Penrose drain/col-
ored tourniquet to make passing the bowel through the transverse
mesocolon defect easier (optional).
7. Mobilization of the common bile duct within the hepatoduodenal
ligament—mobilizing the hepatic flexure of the colon and performing wide
Kocher maneuver should be used liberally to expose the porta hepatis:
002086389.INDD 348 7/14/2014 8:12:29 PM

a. If no cholecystectomy has been performed previously, use the gall-

bladder to identify the cystic duct, and travel proximally toward the
common hepatic–bile duct junction. The gallbladder also provides
a useful “handle” (Fig. 28.2) for a lateral to medial approach when
dissecting the soft tissues of the hepatoduodenal ligament. Typically,
there is a station 12 lymph node just cephalad to the duodenum along
the lateral border of the ligament to mark the location of the distal
common bile duct.
b. Incise the peritoneum anteriorly to expose the bile duct and hepatic
arterial supply. The bile duct will be lateral to the proper hepatic
artery. Be mindful that the right hepatic artery travels anterior to the
common hepatic duct in 10% of patients.
c. Mobilize the ligament further by incising the peritoneum laterally
toward the inferior vena cava to open near the shared border of the
common bile duct and portal vein. This maneuver will allow for less
resistance when dissecting medially to laterally posterior to the bile
duct.
d. Encircle the bile duct with a vessel loop or Penrose drain; use metallic
clips to secure both tails together for retraction purposes.
e. Pass a right-angled instrument posterior to the bile duct, and use
the cautery to divide the bile duct. The cautery can be used to obtain
hemostasis at the 3 and 9 o’clock arterial supple at the cut edge of
the bile duct. The bile duct should bleed briskly; if it does not, then
consider dissecting the bile duct proximally to better perfused tissue.
f. If a biliary endoprosthesis or percutaneous biliary drain is across the
bile duct, either can be used to stent your anastomosis. Otherwise,
if it is cumbersome, stents can be removed as it may make the anas-
tomosis more difficult. If stenting the biliary–enteric anastomosis is
desired, the ideal situation is to place your own 12-Fr biliary stent
across the anastomosis after the posterior row has been completed.
FIGURE 28.2 The gallbladder (if in situ) provides a useful landmark and “handle” to distract
the bile duct laterally.
0002086389.INDD 349 7/14/2014 8:12:30 PM

g. A laparoscopic bulldog clamp is commercially available and can be

used to clamp the proximal cut edge of the bile duct; in some cases,
this may allow for a slight dilatation of a smaller bile duct and mini-
mize biliary soilage of the peritoneum.
h. The distal bile duct should be oversewn or may be closed with a staple
(or clips if small).
8. Biliary–enteric anastomosis:
a. Some surgeons prefer the Trendelenburg position (head down) at the
time of an intracorporeal hand-sewn hepaticojejunostomy to take
tension off of the anastomosis by way of the small bowel mesentery
using gravity to bring the Roux limb closer. Reverse Trendelenburg
position is useful in obese patients by using gravity to retract the
hepatic flexure away from the porta hepatis. The Roux limb is usu-
ally supported with a grasper from one of the assistant port sites in
either scenario.
b. For larger bile ducts, tie two separate 4-0 absorbable sutures together
leaving it double armed. Each suture is cut to a shorter length typically
15 to 20 cm each in length but can be adjusted based on the size of the
bile duct. For small bile ducts, interrupted 4-0 absorbable sutures are
used for the posterior and anterior rows (Fig. 28.3).
c. An enterotomy is created on the anterior aspect (i.e., bowel side clos-
est to the camera) of the antimesenteric border of the Roux limb.
d. A stay suture is placed at 12 o’clock on the bile duct for retraction of
the anterior wall during creation of the posterior row.
e. The suture is introduced into the peritoneum; one arm is passed
outside-in on the bowel and the other on the bile duct (starting at
3 o’clock).
f. The posterior row is created first by suturing inside out on the bowel
then outside-in on the bile duct. Special attention is necessary during
the first bite with the suture as most leaks will occur posteriorly on the
bile duct side after the initial anchoring of the double-armed suture.
g. Upon completion of the posterior row, a bulldog clamp is placed on
the remainder of the posterior row stitch to maintain tension prevent-
ing unraveling of the posterior row.
FIGURE 28.3 The posterior row of absorbable sutures is placed with knots on the inside of
the anastomosis to facilitate visualization of the knots being tied.
002086389.INDD 350 7/14/2014 8:12:31 PM

h. The 12 o’clock stay suture should be removed now to prevent

confusion.
i. Stenting the biliary–enteric anastomosis would occur either now or
midway through completing the anterior row.
j. The anterior row is completed by suturing outside-in on the bowel
then inside out on the bile duct.
k. The anastomosis is completed by tying the two sutures on the outside
at the 9 o’clock position.
l. After completion of the anastomosis, the redundant Roux limb is
pulled back through the mesocolon defect toward the jejunojejunos-
tomy. Furthermore, the limb is secured directly to the peritoneum
at the mesenteric defect with 3-0 nonabsorbable suture to prevent
herniation.
m. Surgical drainage of the anastomosis is by surgeon preference.
POSTOPERATIVE MANAGEMENT, INCLUDING COMPLICATIONS

The majority of the postoperative management centers on bowel function
and the advancement of the patient’s diet. The previously established overall
complication rate for open hepaticojejunostomy is 10% to 20% and mortal-
ity of 1% (Table 28.2). The principal complication for hepaticojejunostomy is
the development of a biliary fistula (early) or stricture (late). The number of
laparoscopic hepaticojejunostomy reports is so low that a true incidence of
these complications is undetermined. A recent systematic review reported
19 studies to date describing a collective experience with true laparo-
scopic choledochoduodenostomy, cholecystojejunostomy, or hepaticojeju-
nostomy for benign or malignant biliary disease (Table 28.2); the authors
reported a biliary fistula or early stricture rate that appears to be greater
than that of the open approach. The corresponding question that presently
has no answer is whether a biliary fistula in the laparoscopic approach
has worse outcomes as compared to the open approach. When comparing
laparoscopic and open reconstruction, it is important to note that many
surgeons with open biliary reconstruction currently do not internally stent
or externally drain this anastomosis. Routine drainage of laparoscopic
biliary–enteric anastomosis may be prudent as the incidence of early biliary
fistula may be slightly higher using the laparoscopic approach. Although
not common, it is possible to see bile in the surgical drain on postoperative
day 1. Nevertheless, in limited experience with these biliary fistulas, they
resolve in 2 to 3 days later and the majority prior to the date of discharge.
TABLE
Comparison of Laparoscopic and Open Biliary–Enteric
28.2 Reconstruction
Laparoscopic Open
Patients Reviewa (N = 89) Historical Comparison
Conversion to open 2 (2%) N/A
Early patency rate (90 d) 85 (96%) >90%
Mortality 5 (6 %) 1%
Morbidity 11 (12%) 15%
Biliary fistula 2 (2%) 1%
Early biliary stricture (1 y) 2 (2%) 1%
Long-term patency rate (>20 y) N/A 70%–90%
a
Toumi Z, Aljarabah M, Ammori B. Role of laparoscopic approach to biliary bypass for benign and
malignant disease: a systematic review. Surg Endosc 2011;25:2105–2116.
0002086389.INDD 351 7/14/2014 8:12:31 PM

If a prolonged fistula is seen, then the surgical drain is left in place at the
time of discharge, and the suction bulb is changed to a gravity drainage
bile bag. Prolonged fistulas are managed as an outpatient as long as the
patient is asymptomatic, and at each clinic visit, the drain is retracted a
short distance (2 to 3 cm) away from the fistula and secured to the skin.
As long as the surgical drain is in place, rarely does a biliary fistula require
intervention or lead to biliary sepsis. A high-volume (i.e., >100 to 200 mL)
biliary fistula demands more thorough interrogation and more aggressive
treatment.
Further complications include a jejunojejunostomy leak (<1%), which
is exceedingly rare and does not typically require external drainage. The
likely cause of this complication is due to ischemia of the bowel by inat-
tentive division of the arterial supply with the tissue-sealing device within
the mesentery while creating the Roux limb. If the bowel appears dusky or
blue, it is best to repeat the anastomosis instead of hoping it improves. Some
surgeons routinely create BP and Roux limbs 50 cm each, which gives extra
length of bowel in case one needs to repeat anastomosis either at the time of
the index procedure or during a take-back/revision. Internal hernia is more
likely in larger patients and those where the mesocolon and jejunojejunos-
tomy mesenteric defects were not closed. In spite of suture closure of defects,
patients still likely have a 5% to 10% lifetime risk of developing a symptom-
atic internal hernia. These hernias present as a bowel obstruction, and any
patient with a Roux limb should have an increased suspicion for internal
hernia and a low threshold for operative exploration especially considering
nonoperative management rarely resolves this type of bowel obstruction.
The development of port site hernia is rare for 5-mm port sites and less than
1% for 12-mm port sites. Port sites do not require fascial closure if they are
5 mm; some surgeons do not close 12-mm port sites above the umbilicus
since they are “protected” by the liver and stomach. A prudent approach is
to close all fascial defects at 12-mm port sites either conventionally in thin
patients or laparoscopically with the suture passer in obese patients.
OUTCOMES/FOLLOW-UP
The primary long-term outcome of interest following biliary–enteric recon-
struction is patency rate. Open series report a 70% to 90% patency rate
two decades following Roux-en-Y hepaticojejunostomy (Table 28.2). Data
for laparoscopic biliary reconstruction are scant; no long-term data exist
documenting true patency rates.
Follow-up varies widely, yet most surgeons see their patients 3 to 4
weeks postoperatively with a history and physical exam and laboratory
studies (CBC, CMP) to determine if the patient has any signs of biliary fis-
tula or stricture and to ensure adequate biliary drainage. A more thorough
approach is to obtain a right upper quadrant ultrasound at 12 weeks to
determine the presence of any dilated intrahepatic ducts or postoperative
fluid collection in the gallbladder fossa, followed by a 99mTc choletech scan
in nuclear medicine at 24 weeks to document filling of the Roux limb with
radiotracer from the liver. In cases where a biliary fistula or stricture is iden-
tified in the postoperative period, cholangiography is indicated; the percu-
taneous approach is favored over double-balloon enteroscopy via the Roux
limb since intervention is more accessible via the percutaneous approach.
CONCLUSION
Minimally invasive biliary surgery has not progressed as rapidly as mini-
mally invasive liver or pancreatic surgery. The principal hurdle is the tech-
nical challenge and paramount reliance of intracorporeal suture at difficult
002086389.INDD 352 7/14/2014 8:12:31 PM

angles in the “faraway” porta hepatis where retraction and exposure are
technically challenging. Optimal outcomes are achieved by proper patient
selection as well as surgeon experience both with biliary pathology and
with complex laparoscopic skills.
Suggested Readings
Bismuth H. Postoperative strictures of the bile duct. In: Blumgard LH, ed. The biliary
tract. Edinburgh, UK: Churchill Livingstone, 1982:209–218.
Branum G, Schmitt C, Baillie J, et al. Management of major biliary complications after
laparoscopic cholecystectomy. Ann Surg 1993;217:532–540.
Chowbey PK, Katrak MP, Sharma A, et al. Complete laparoscopic management of cho-
ledochal cyst: report of 2 cases. J Laparoendosc Adv Surg Tech A 2002;12:219–223.
Chowbey PK, Soni V, Sharma A, et al. Laparoscopic hepaticojejunostomy for biliary
strictures. The experience of 10 patients. Surg Endosc 2005;19:273–279.
Lillemoe KD, Melton GB, Cameron JL, et al. Postoperative bile duct strictures: man-
agement and outcome in the 1990s, Ann Surg 2000;232:430–441.
Shimura H, Tanaka M, Shimizu S, et al. Laparoscopic treatment of congenital chole-
dochal cysts. Surg Endosc 1998;12:1268–1271.
Stewart L, Way LW. Bile duct injuries during laparoscopic cholecystectomy, Arch Surg
1995;130:1123–1128.
Strasberg SM, Hertl M, Soper NJ. An analysis of the problem of biliary injury during
laparoscopic cholecystectomy, J Am Coll Surg 1995;180:101–125.
Toumi Z, Aljarabah M, Ammori B. Role of laparoscopic approach to biliary bypass
for benign and malignant disease: a systematic review. Surg Endosc 2011;
25:2105–2116.
0002086389.INDD 353 7/14/2014 8:12:31 PM

29 Bile Leak and Bile Duct Injury
Jordan P. Bloom and Keith D. Lillemoe
INTRODUCTION
A biliary leak is defined as leakage of bile from any site in the biliary tree
including the bile duct, cystic duct, liver, or gallbladder. In the United States,
a bile duct injury (BDI) is most commonly the result of iatrogenic injury
during cholecystectomy. Cholecystectomy is one of the most commonly
performed abdominal operations with over 750,000 performed yearly in the
United States. Over 90% of these operations are being performed laparo-
scopically. The overall incidence of bile duct injuries during laparoscopic
cholecystectomy is reported to range from 0.2% to 0.5%. This incidence is
higher than the reported incidence in the “open cholecystectomy” era (0.1%
to 0.2%). However, concerns currently exist that surgeons trained almost
exclusively in laparoscopic cholecystectomy may lack adequate training
and experience in the difficult open cholecystectomy. Therefore, the inci-
dence of BDIs during open or converted cholecystectomy may increase in
both frequency and severity.
A major challenge associated with biliary injuries is that such inju-
ries are recognized at the time of laparoscopic cholecystectomy less than
40% of the time. Thus, most injuries present in the early postoperative
period often with a biliary leak or jaundice. The therapeutic options, as
well as the condition of the patient varies greatly based on the timing
and mode of presentation, but in most cases, there should be no rush to
return the patient to the operating room without proper evaluation and
often nonsurgical intervention. Timely diagnosis and appropriate ther-
apy are, however, necessary to promote optimal clinical outcome. The
diagnosis and management of biliary leaks have been greatly facilitated
by recent advances in imaging as well as percutaneous and endoscopic
interventional techniques. The optimal management of such patients is
best provided in centers that offer a multidisciplinary team approach.
The basic principles of management of almost all biliary leaks includes
control of the biliary leak by drainage of the biliary tree, control of the
intra-abdominal sepsis, definition of biliary anatomy, and, when appro-
priate, definitive repair. This chapter reviews the etiology, clinical presen-
tation and diagnosis, management, and outcomes with respect to BDI
and biliary leak.
ETIOLOGY OF BILIARY LEAK

A bile leak is almost always the result of iatrogenic injury or a complication
of surgical, percutaneous, or endoscopic procedures (Table 29.1).
There are a number of classifications of biliary leaks and injuries. In
the United States, the Strasberg classifications have become the most com-
monly employed classification (Fig. 29.1). The majority of bile leaks are a
result of Strasberg type A injuries.
354
002141491.INDD 354 7/14/2014 8:15:24 PM

Chapter 29 / Bile Leak and Bile Duct Injury 355
TABLE
29.1 Etiology of Bile Leak
Abdominal operations
• Cholecystectomy
∞ Major BDI
∞ Minor bile leak (cystic duct, duct of Luschka)
• Pancreaticobiliary resection
• Biliary reconstruction
• Hepatic resection
• Hepatic transplantation
• Gastroduodenal surgery
Percutaneous interventions
• Transhepatic cholangiography/drainage (PTC)
• Stricture dilation
• Radiofrequency tumor ablation
• Embolization
• Liver biopsy
Endoscopic retrograde cholangiography perforation
Trauma (blunt and penetrating)
Biliary leaks can also be classified as either a minor bile leak or a major
biliary ductal injury. Minor bile leaks are typically a result of cystic duct
stump leak (60% to 75%) or duct of Luschka injuries (10% to 20%). Ducts of
Luschka are accessory bile ducts located in the gallbladder bed that do not
communicate with the lumen of the gallbladder. While they do not drain
any liver parenchyma, they are a common source of minor bile leak after
cholecystectomy. Major biliary ductal injury is most often an injury to the
common bile duct (CBD) or common hepatic duct and includes avulsion,
ligation, and transection. A bile leak can also be classified based on the vol-
ume of drainage. Low-output drainage is considered less than 300 mL of
bile per day with high-output drainage being greater than 300 mL/day.
While cholecystectomy is the commonest cause of bile leak, it is
important to remember that bile leakage can complicate any procedure
that has a bilioenteric anastomosis. The exact incidence of biliary leak after
these procedures is difficult to determine due to the wide variety of possible
procedures; however, it is likely in the range of 2% to 5%. One of the great-
est challenges in the diagnosis of leak after a bilioenteric reconstruction is
that typically a jejunal Roux-en-Y limb is involved, which creates an access
problem for traditional endoscopic diagnostic techniques.
In the last 15 years, there has been a marked increase in the complex-
ity of liver surgery with a concomitant increase in the bile leak rate. The
incidence of bile leak after hepatic resection is between 2% and 10% with
most of the leaks being minor. A recent prospective analysis of 2,628 hepatic
resections found that while liver-related complication rates remained sta-
ble since 1997, bile leak rates increased (5.9% from 3.7%). These authors
reported that independent predictors of bile leak included bile duct resec-
tion, extended hepatectomy, repeat hepatectomy, en bloc diaphragmatic
resection, and intraoperative transfusion. Since it is known that bile leak
is associated with both increased hospital stay and mortality, novel tech-
niques to minimize leak after hepatic resection have been introduced with
numerous ongoing trials of new methodologies to accomplish this goal.
Bile leaks continue to be an uncommon complication of liver trans-
plantation. Despite great improvements in surgical technique, biliary
0002141491.INDD 355 7/14/2014 8:15:24 PM

FIGURE 29.1 Strasberg classification. Type A: Injury to the cystic duct or from minor hepatic
ducts draining the liver bed. Type B: Occlusion of biliary tree, commonly aberrant right hepatic
duct(s). Type C: Transection without ligation of aberrant right hepatic duct(s). Type D: Lateral
injury to a major bile duct. Type E (1–5): Injury to the main hepatic duct; classified according
to level of injury (i.e., below). E1 (Bismuth type 1): Injury more than 2 cm from confluence;
E2 (Bismuth type 2): Injury less than 2 cm from confluence; E3 (Bismuth type 3): Injury at
the confluence; confluence intact; E4 (Bismuth type 4): Destruction of the biliary confluence;
E5 (Bismuth type 5): Injury to aberrant right hepatic duct. (Reproduced from Winslow ER,
Fialkowski EA, Linehan DC, et al. “Sideways”: results of repair of biliary injuries using a policy
of side-to-side hepatico-jejunostomy. Ann Surg 2009;249(3):426–434, with permission.)
002141491.INDD 356 7/14/2014 8:15:25 PM

complications are a common source of morbidity, which result in the need

for intervention including percutaneous, endoscopic, or even surgical
procedures. The incidence of biliary complications currently ranges from
5% to 25% with biliary strictures (9% to 12%) and leaks (5% to 10%) being
the most common complications. Bile leak after liver transplant is almost
always related to the biliary reconstruction and is complicated by the need
for immunosuppression in the patient population.
A bile leak after blunt and penetrating abdominal trauma involving
the liver is seen in 5% to 10% of reported cases and can occur after either
operative or nonoperative management.
ETIOLOGY OF BILE DUCT INJURY

In the United States, BDIs are most commonly the result of injury during
laparoscopic cholecystectomy. CBD injury during cholecystectomy can
occur due to either a technical error or more commonly misidentification
of the ductal anatomy. The Stewart-Way classification is a clear and con-
cise method of classifying for the type of CBD injury based on the specific
mechanism of injury (Table 29.2). The most common mechanism is a class
III injury where the CBD is mistaken for the cystic duct with the error being
unrecognized at the time of injury (Fig. 29.2). The duct is then transected
and a segment usually resected. In the case series used to generate the
Stewart-Way classification scheme, class III injuries were observed in 60%
of the cases.
Contributing factors to BDI include acute or chronic inflammation in
the triangle of Calot, a short cystic duct, excessive cephalad retraction on
the gallbladder fundus, and insufficient or excessive lateral retraction of the
gallbladder infundibulum. Other factors that can increase the likelihood of
injury include use of an end-viewing scope, excessive use of cautery, physi-
cian inexperience, and aberrant biliary anatomy.
The “critical view of safety” (CVS) was first described in 1995 by
Strasberg et al. This method has been adopted increasingly by surgeons
around the world for performance of laparoscopic cholecystectomy
(Fig. 29.3). The CVS has three requirements. First, the triangle of Calot must
be cleared of fat and fibrous tissue. The second requirement is that the low-
est part of the gallbladder be separated from the cystic plate, the flat fibrous
surface to which the nonperitonealized side of the gallbladder is attached.
The third requirement is that two structures, and only two, should be seen
entering the gallbladder. Once these three criteria have been fulfilled, CVS
TABLE
Mechanism of Common Bile Duct Injury (Stewart-Way
29.2
Class I
Classification)
CBD mistaken for cystic duct but recognized
Cholangiogram incision in cystic duct extended into CBD
Class II Lateral damage to the CHD from cautery or clips placed on duct
Associated bleeding, poor visibility
Class III CBD mistaken for cystic duct, not recognized
CBD, CHD, R, L. hepatic ducts transected and/or resected
Class IV RHD mistaken for cystic duct, RHA mistaken for cystic artery,
RHD and RHA transected
Lateral damage to the RHD from cautery or clips placed on duct
CBD, common bile duct; CHD, common hepatic duct; L, left; R, right; RHA, right hepatic artery; RHD,
right hepatic duct.
Reprinted from Way LW, et al. Causes and prevention of laparoscopic bile duct injuries. Ann Surg
2003;237(4):460, with permission.
0002141491.INDD 357 7/14/2014 8:15:25 PM

Accidentally
divided
hepatic
ducts
Cystic duct
Common hepatic duct
Common
bile duct
FIGURE 29.2 Classic laparoscopic BDI. The CBD is mistaken for the cystic duct and tran-
sected. A variable extent of the extrahepatic biliary tree is resected with the gallbladder.
The right hepatic artery, in background, is also often injured. (Adapted from Branum G,
Schmidt C, Baillie J, et al. Management of major biliary complications after laparoscopic
cholecystectomy. Ann Surg 1993;217:532.)
FIGURE 29.3 Operative photo demonstrating the CVS with extensive separation of the
lower gallbladder from the cystic plate. The cystic duct and cystic artery are clearly defined.
002141491.INDD 358 7/14/2014 8:15:29 PM

has been attained. There is no level I evidence that CVS reduces BDI, and
there likely never will be as the incidence of injury is so low that it would
be difficult to adequately power a prospective randomized trial. There are
numerous case series, however, which show a reduction in the incidence
of BDI when the CVS is attained, such that it is widely considered to be the
“safest” technique for performing laparoscopic cholecystectomy.
The role of intraoperative cholangiography (IOC) in preventing BDI
during laparoscopic cholecystectomy is controversial. Individual insti-
tutional series have failed to demonstrate that either routine or selective
cholangiography affects the incidence of BDI. Furthermore, large databases
have provided mixed results as to whether IOC is an effective strategy to
reduce the incidence of BDI.
The clinical presentation of postoperative bile leak is variable. If surgical
drains are in place, a leak is obvious and easily detectable. Assuming there
are not drains in place or the drains do not communicate with the area of
leakage, bile can leak freely into the peritoneal cavity or it can loculate as a
collection. The presentation can be subtle ranging from nonspecific com-
plaints such as abdominal discomfort, nausea, and low-grade temperate to
severe abdominal pain and/or a “septic” appearance due to bile peritonitis.
Therefore, clinicians should have a high index of suspicion for bile leak in
patients who have undergone right upper quadrant surgical procedures and
who have any postoperative deviation from the projected clinical course.
DIAGNOSIS
There are numerous diagnostic modalities to consider in the evaluation of
a potential bile leak. After history and physical exam, the first step in the
workup of suspected bile leak should be basic laboratory studies. While
laboratory values may be totally normal, a mild elevation in the serum
total bilirubin, often due to bile absorption from the peritoneum, may be
seen. Other liver function tests are often normal. Leukocytosis representing
inflammation or infection is common.
Following laboratory evaluation, there are numerous imaging modali-
ties to consider.
Ultrasonography is an inexpensive and noninvasive imaging tech-
nique that may be a useful first step in identifying an intra-abdominal fluid
collection; however, numerous studies have concluded that the sensitiv-
ity of ultrasound for detection of bile leak is significantly inferior to other
modalities. Abdominal axial imaging with CT provides a more detailed
view of the pertinent structures as well as fluid collections or bile ascites
(Fig. 29.4). Although the exact nature of the fluid as being bile may be uncon-
firmed, such collections certainly warrant suspicion and further workup.
The gold standard for diagnosis of biliary leakage is cholangiography.
There are numerous ways that a cholangiogram can be achieved in the
postoperative patient. If the patient has operative drains in place, contrast
can be injected in a retrograde fashion that may demonstrate the point of
leakage and biliary anatomy. Operatively placed T tubes or transhepatic
stents also provide direct access to the biliary tree. In patients who have
no access to the biliary tree, cholangiography can be performed either
antegrade via percutaneous transhepatic puncture (PTC) or, more com-
monly, retrograde via endoscopic retrograde cholangiopancreatography
(ERCP). Both of these techniques require skilled operators and confer some
degree of risk. Assuming the patient has native anatomy, the first option
would be ERCP. Although ERCP is generally well tolerated by patients and
0002141491.INDD 359 7/14/2014 8:15:29 PM

FIGURE 29.4 Large bile duct collection (biloma) occurring after BDI.
allows therapeutic intervention (stent placement and/or sphincterotomy)

at the time of diagnosis, a recent large European study reported an overall
complication rate of 10%. Post-ERCP pancreatitis occurred in 4.2%, bleed-
ing in 3.6% (0.4% clinically relevant), cholangitis in 1.4%, cardiopulmonary
complications in 1.2%, perforation in 0.6%, and procedure-related deaths in
0.1% of procedures. The limitation of ERCP is that in cases with disruption
of the normal biliary anatomy, either through biliary transection or anasto-
mosis, an ERCP will not demonstrate the proximal biliary tree and therefore
not define the location of the leak or the anatomy necessary for reconstruc-
tion (Fig. 29.5). PTC may therefore be the only access to the biliary tree in
certain situations (Fig. 29.6). PTC is another safe technique with a major
complication rate of 2% to 10% (sepsis, cholangitis, bile leak, hemorrhage,
or pneumothorax). After gaining access to the biliary tree, PTC also offers
the potential for therapeutic intervention.
A noninvasive modality for imaging the biliary tree that might be
useful in patients who are clinically stable is magnetic resonance cholan-
giopancreatography (MRCP). Recent developments in imaging technology
have enabled MRCP to replace invasive alternatives for the pre- and post-
operative assessment of biliary disease. MRCP enables rapid, noninvasive
evaluation of both the biliary tree and pancreatic duct without the use of
intravenous contrast media. Attention has recently focused on the utility in
evaluation of the biliary system in suspected biliary injuries to help select
the most appropriate next step if any (Fig. 29.7). The major limitation to
MR-based biliary imaging is that there is no potential for therapeutic inter-
vention. Other limitations include high cost, patient size restrictions, and
the requirement of lying still for a prolonged period of time, which might be
difficult in the acutely ill patient.
Finally, technetium-99m–labeled hydroxy iminodiacetic acid (HIDA)
is commonly used as initial, noninvasive imaging modality for a sus-
pected biliary leak. HIDA scintigraphy is a dynamic study in which an
ongoing bile leak may be detected. However, it provides suboptimal
anatomic detail. A biliary leak will be demonstrated by extravasation
002141491.INDD 360 7/14/2014 8:15:29 PM

FIGURE 29.5 An ERCP performed in a patient with a major BDI. Contrast does not fill the
proximal biliary tree due to operatively placed clips.
of radionuclide from the biliary system; however, these studies always

require further investigation to define the true nature of the leak and to
perform biliary drainage.
MANAGEMENT OF BILE LEAKS AND INJURIES

Management of Minor Bile Leaks after Cholecystectomy
Minor bile leakage following laparoscopic cholecystectomy is most
commonly due to cystic duct stump leak (60% to 75%) or from a duct of
Luschka (10% to 20%). The same basic management strategy applies to
any minor bile leak, which includes decompression of the biliary tree,
usually via endoscopic means, and drainage of the intra-abdominal col-
lections to control sepsis. The most effective method of decompressing
the biliary tree is lowering the pressure gradient across the sphincter
of Oddi. This may be accomplished endoscopically via sphincterotomy
with or without transampullary stenting. These techniques decompress
the biliary tree and allow bile to take the path of least resistance into
the duodenum, which usually allows healing of the upstream pathology.
The endoscopic route with or without stenting is preferred as the endo-
scopic approach is easier, particularly as the intrahepatic bile ducts are
often not dilated, and it is better tolerated by patients. Percutaneous
transhepatic drainage is another option although is seldom necessary
0002141491.INDD 361 7/14/2014 8:15:30 PM

FIGURE 29.6 Percutaneous transhepatic cholangiogram demonstrates complete obstruc-

tion of the biliary tree at the site of bile duct transection.
since most minor bile leaks following cholecystectomy involve intact

biliary anatomy. An increasingly common indication for PTC is a bile
leak or injury following cholecystectomy after gastric bypass proce-
dures. While the success in healing minor bile leaks using various meth-
ods have not been compared to each other using randomized studies,
most series quote success rates of nonoperative decompression greater
than 90%. A recent study by Pitt et al. analyzed 528 patients with biliary
leak or duct injury managed by endoscopists, interventional radiolo-
gists, and surgeons. They concluded that bile leaks from the cystic duct
or a duct of Luschka were managed almost exclusively by endoscopists
with excellent results.
Management of Bile Leak after Liver Resection

Bile leak after liver resection continues to be a fairly common problem
complicating 5% to 10% of cases. The majority of cases are minor leaks
from the transected liver parenchymal surface that do not cause signifi-
cant morbidity. However, major biliary complications may occur and
cause significant morbidity and even mortality. Such injuries include
injury to the common bile or hepatic duct, from the stump of a ligated
main biliary trunk or from a biliary–enteric anastomosis. Bile may also
leak from an immature T tube or transhepatic stent tract. Many investiga-
tors have tried to determine risk factors for bile leak after liver surgery.
There does not seem to be a significant correlation with the preoperative
condition of the patient, liver pathology, or technique used to divide the
hepatic parenchyma. Some studies have suggested that longer operative
times and the removal of Couinaud segment 4 may be independent risk
factors for bile leak.
002141491.INDD 362 7/14/2014 8:15:30 PM

FIGURE 29.7 A: MRCP demonstrates intact biliary anatomy with a small leak from the
cystic duct. In such a case, a therapeutic ERCP would be indicated. B: MRCP demonstrates
complete transection of the common hepatic duct. In this case, ERCP would be of no value,
and a percutaneous approach would be necessary to access the biliary tree should drain-
age be necessary. Alternatively, if appropriate, the patient could be taken directly to the
operating room for reconstruction as the anatomy is nicely defined.
0002141491.INDD 363 7/14/2014 8:15:31 PM

Bile leaks after hepatectomy may present in the early postopera-

tive period (days 2 to 4) with bile detectable in surgically placed drains.
However, since drains are not routinely used by many liver surgeons, there
may be significant delay in recognition of bile leak in patients who do not
have drains in place. These patients often present 6 to 8 days postopera-
tively with one or more of the following: abdominal pain, low-grade fever,
leukocytosis, hyperbilirubinemia, ileus, or protracted hiccups. Rarely, a bili-
ary leak may present as a severe septic or hemorrhagic event.
The workup of bile leak after hepatic resection is similar to that for
postcholecystectomy bile leak. The most likely source of bile leak after
hepatic resection is leakage from the parenchyma, which tends to be low
volume and will often resolve with external drainage of the collection and
antibiotics. If the bile leak persists, decompression of the biliary tree usu-
ally via endoscopic techniques is appropriate if native anatomy is present.
Management of Bile Leak after Liver Transplantation

Biliary complications continue to be a major source of morbidity in hepatic
transplantation. The major biliary complications are bile leak and stricture,
occurring in 5% to 25% of transplants. While there are exceptions, most bile
leakage after transplant occurs at the biliary anastomosis. Currently, the
most common anastomotic techniques are direct choledochocholedochos-
tomy (end-to-end or side-to-side) with or without T-tube stenting or Roux-
en-Y choledochojejunostomy with no clear difference in the incidence of
biliary leak.
Bile leakage in transplant patients presents similar to other patient
populations discussed earlier. However, due to a suppressed immune
response, these patients may not mount the same magnitude of response
mostly with respect to fever, leukocytosis, and abdominal pain. As such,
a high awareness of the potential for bile leak in the early postoperative
period is important.
Treatment strategies for bile leak after liver transplant are essentially
the same as discussed for other types of bile leaks. The use of nonopera-
tive strategies has nearly obviated the need to return to the operating room
for anastomotic revision except in the most drastic situations. In patients
with direct choledochocholedochostomy, ERCP with sphincterotomy and
stenting is the treatment of choice with a frequency of success approaching
90%. However, in patients with Roux-en-Y choledochojejunostomy, ERCP
is more difficult and requires a highly skilled interventional biliary endos-
copist. Percutaneous transhepatic stenting of the biliary anastomosis may
also be technically difficult in patients who do not have intrahepatic biliary
ductal dilation and often the transplant surgeon wishes to avoid this route.
Management of Bile Leak after Trauma

The liver is the most frequently injured abdominal organ in both penetrat-
ing and blunt trauma. Nonoperative management has become well estab-
lished as the standard of care for treatment of patients who have blunt liver
injury. This development has resulted, in large part, from the widespread
use of abdominal CT, which allows more precise definition of the extent of
injury, and from the increased availability and application of angiography/
embolization to control hemorrhage in these patients. Currently, 60% to
85% of all patients who have blunt hepatic trauma are managed success-
fully without operation. As a result of this paradigm change, it is difficult
to estimate accurately the true incidence of bile leak after hepatic trauma.
Biloma or bile ascites has been reported in as few as 0.5% to as many as
20% of all patients suffering blunt liver injury. In the most comprehensive
study of bile leak after blunt liver trauma, Wahl et al. reported 24 bile leaks
002141491.INDD 364 7/14/2014 8:15:31 PM

in 258 patients for an overall incidence of 9%. Several reviews have found an
increased incidence of bile leak in patients who have more complex injuries
and in those undergoing angiographic embolization, with an overall inci-
dence of 30% to 40% in these populations.
The diagnosis of bile leak may prove particularly difficult in patients
who have blunt trauma managed nonoperatively. Associated injuries may
mask the typical symptoms and physical signs of bile peritonitis. Conversely,
nonspecific symptoms or abnormal laboratory values may be attributed to
other known or suspected injuries. Routine imaging studies (CT and ultra-
sound) can demonstrate intra-abdominal and perihepatic fluid collections.
Such screening of patients managed nonoperatively at high risk for bile leak
(those who have complex injuries or are undergoing angiographic emboli-
zation) may prove particularly useful in the trauma population.
As may be expected, with the paucity of data in the literature, no clear
treatment strategy has evolved for managing bile leaks after liver injury.
Several authors have reported success with simply percutaneous drain-
age of bile collections. Like with other causes of bile leak, draining the bili-
ary tree using ERCP +/− sphincterotomy, or endoscopic stent placement,
are options for persistent leaks. In some cases with major ductal injuries,
operative management including reconstruction or even resection may be
necessary.
Recently, laparoscopy with peritoneal lavage and operative drain
placement has been suggested as a useful modality for treating patients
who have hepatic trauma. Wahl noted that patients who were diagnosed as
having a bile leak and were treated early by laparoscopic peritoneal lavage
made more rapid recovery and had earlier hospital discharge than those
diagnosed and treated later in the postinjury course. This finding prompted
them to recommend early (postinjury day 3 or 4) screening for bile leak with
HIDA scan in patients at high risk for bile leak, with laparoscopic peritoneal
lavage applied to patients who had positive scans. Although early laparo-
scopic peritoneal lavage has not been studied in a randomized fashion, it
may hold promise for this specific group of patients.
Most penetrating liver injuries continue to be managed operatively.
Furthermore, a small percentage of patients who have blunt liver injury still
require operative intervention. Diagnosis and management of bile leaks
in these cases is similar to the management of bile leaks following other
operative procedures on the liver or biliary tree. Biliary leak in these cases
generally is suggested by the presence of bile in an operatively placed drain.
Clinical findings of low-grade fevers, tachycardia, ileus, increased serum
bilirubin, or leukocytosis may prompt abdominal CT and reveal an und-
rained fluid collection. External drainage of bile collections (by CT-guided
percutaneous drainage), control of sepsis with tailored antibiotic treat-
ment, and subsequent endoscopic or percutaneous transhepatic decom-
pression of the biliary tree generally leads to resolution of bile leak in these
situations.
Management of Major Bile Duct Injury

Most major bile duct injuries in the United States occur during laparo-
scopic cholecystectomy. Optimal management depends on timing of recog-
nition of the injury. Most of these injuries (60% to 70%) go unrecognized at
the time of laparoscopic cholecystectomy. Figure 29.8 shows the suggested
management of injuries associated with laparoscopic cholecystectomy.
If the injury is recognized at the time of cholecystectomy, the first
step is to determine if the operating surgeon is technically capable of per-
forming the necessary repair. If he or she is not, we suggest consulting
another surgeon at the original hospital who is experienced in biliary tract
0002141491.INDD 365 7/14/2014 8:15:31 PM

LAPAROSCOPIC CHOLECYSTECTOMY
MAJOR BILE DUCT INJURY
RECOGNIZED AT TIME OF RECOGNIZED IN THE EARLY

LAPAROSCOPIC CHOLECYSTECTOMY (30%) POSTOPERATIVE PERIOD
PRIMARY REPAIR TRANSFER TO CT SCAN

HEPATOBILIARY CENTER (DRAIN COLLECTIONS)
END-TO-END HEPATICOJEJUNOSTOMY
REPAIR (PREFERRED IN MOST SETTINGS) PERCUTANEOUS TRANSHEPATIC
CHOLANGIOGRAPHY WITH BILIARY
DRAINAGE
DEFINE ANATOMY/CONTROL BILE

LEAK
4 to 6 WEEKS
DEFINITIVE REPAIR
HEPATICOJEJUNOSTOMY
FIGURE 29.8 Algorithm for diagnosis and management of BDI associated with laparo-
scopic cholecystectomy.
r econstruction. If there is no one available who can perform such a repair,

the patient should be transferred to a tertiary care center for further man-
agement. Once an injury is recognized, it is usually best to avoid further
dissection or clipping of the proximal ducts. It is, however, mandatory to
create a controlled biliary fistula prior to closing. This can be easily accom-
plished by placing a small drainage catheter retrograde into the biliary duc-
tal system as well as placing a closed suction drain into the gallbladder bed.
At this point, the patient should be closed and transferred to a center with
skilled multidisciplinary teams to manage the injury.
A few specific scenarios deserve further discussion. One such situa-
tion is the placement of a clip across the CBD without cutting or dividing
the duct (Fig. 29.9). In such cases, the clip can be removed, the gallbladder is
removed, and no further treatment is necessary. If a small choledochotomy
has been made to perform a cholangiogram, it should be closed with simple
fine absorbable sutures and a drain should be placed in the gallbladder bed.
The procedures can be performed laparoscopically; however, conversion
to open operation is advisable for most surgeons. The patient is at risk to
develop a late stricture, but in most cases, endoscopic balloon dilation and
stenting will result in successful management.
The other situation in which minimal intervention is appropriate is
when a minor bile duct is transected. In such cases, if the duct is less than
4 mm in diameter and cholangiography demonstrates only subsegmental
liver drainage with the remainder of biliary anatomy intact, simple ligation
is preferable than an attempt at reconstruction.
If a major duct injury is suspected and the surgeon is capable of repair-
ing the injured duct, the first step is to convert to an open procedure and
use gross anatomic identification combined with cholangiography to bet-
ter define the nature of the injury. As discussed above, if the combination
of exploration and cholangiography reveal an injured duct less than 4 mm,
which drains subsegmental hepatic parenchyma, simple ligation is accept-
able. If the injured duct is larger than 4 mm, it likely drains a larger area of
hepatic parenchyma and will typically require operative repair.
002141491.INDD 366 7/14/2014 8:15:32 PM

FIGURE 29.9 Intraoperative cholangiogram demonstrating a normal distal bile duct with
drainage into the duodenum. The proximal biliary tree cannot be visualized due to a clip
placed across the duct. Such a cholangiogram should alert the surgeon that a major BDI will
occur unless steps are taken to redefine anatomy before transecting the duct.
If the injury is to the common hepatic or CBD, it should be repaired

immediately. The two most common methods of repair are primary end-to-
end repair or a bilioenteric anastomosis, typically a Roux-en-Y hepaticoje-
junostomy. If a primary end-to-end repair is to be undertaken, the injured
segment of the bile duct should be less than 1 cm and the ends must be able
to be approximated without tension. This is usually accomplished by per-
forming a generous Kocher maneuver (mobilizing the duodenum by taking
down the lateral attachments from the retroperitoneum). After the Kocher
0002141491.INDD 367 7/14/2014 8:15:32 PM

Hepaticojejunostomy
Silastic
biliary stent
Silastic
biliary stent
in jejunum
End-to-side
Transverse jejunostomy
colon
FIGURE 29.10 Schematic representation of a Roux-en-Y hepaticojejunostomy with trans-

anastomotic stent to repair a BDI. (From Mulholland MW, Lillemoe KD, Doherty GM,
et al., eds. Greenfield’s surgery: scientific principles and practice, 4th ed. Philadelphia, PA:
Lippincott Williams & Wilkins, 2006.)
maneuver, there is typically enough length to facilitate approximation of

the injured ends. The anastomosis should be decompressed with a T tube
placed either above or below the suture line. If the end-to-end repair is con-
traindicated, the distal bile duct should be oversewn and repair undertaken
with a bilioenteric anastomosis.
The long-term results of end-to-end repair have often been criticized
for having a high failure rate with late stricture occurring in over half of
reported cases. However, by keeping normal biliary anatomy intact, endo-
scopic balloon dilatation with stenting is an option for treatment of such
strictures. The results for such treatment are excellent and spare the patient
the need for percutaneous procedures that are often painful and associ-
ated with the need for long-term external stents, which are unpopular with
patients.
Most major bile duct injuries, however, will require a bilioenteric anas-
tomosis with a Roux-en-Y limb of jejunum. The jejunum is divided about
25 cm from the ligament of Treitz. A 60-cm tension-free Roux limb is pre-
pared in standard fashion with a stapled functional end-to-side enteroen-
terostomy. The stapled end of the jejunum is then pulled up through a rent
made in the transverse mesocolon to the right of the middle colic vessels.
Reconstruction is performed using a hand-sewn end-to-side anastomosis
ensuring that the mucosa of the bile duct is brought into contact with the
mucosa of the jejunum (Fig. 29.10). Many hepatobiliary surgeons believe
that external biliary drainage preferably by an operatively placed transhe-
patic stent is necessary in all cases of biliary reconstruction after BDI. This
catheter provides decompression of the biliary tree as well as allows rec-
ognition of and protection against the consequences of an anastomotic
002141491.INDD 368 7/14/2014 8:15:33 PM

FIGURE 29.11 Postoperative cholangiogram after immediate reconstruction of a major

BDI.
leak. This can be easily accomplished intraoperatively by placing a silastic

stent retrograde through the hepatic parenchyma and externalized prior to
completing the anastomosis. Another benefit of the catheter is the ability
to easily study the repair and biliary anatomy postoperatively with cholan-
giography (Fig. 29.11).
Unfortunately, most patients who sustain a major BDI are not recog-
nized during cholecystectomy. These patients usually present with pain,
nausea, vomiting, bloating, and fever early in their postoperative course
(days 2 to 5). These symptoms necessitate further diagnostic evaluation as
described previously. A CT should be performed to evaluate for bile ascites
or a bile collection. Such studies should also include arterial phase contrast,
as up to 20% of major BDI may have an associated hepatic arterial injury.
Assuming there is evidence of bile leak, decompression and external drain-
age of the biliary system is mandated. In cases in which there is evidence of
a bile duct transection, the only reliable route for biliary drainage is using
a percutaneous transhepatic approach. If there are drainable collections,
these should also be drained by interventional radiology (IR). Although
most patients with a bile leak do not have “infected bile,” broad-spectrum
antibiotics to cover typical biliary pathogens should be started as well as
used to provide prophylaxis for various procedures. Once the leak is con-
trolled via external drainage (often allowing operatively placed or IR-placed
drains to be removed), and the biliary system is decompressed, there is no
immediate rush to repair. It is clearly beneficial to allow resolution of the
inflammatory process prior to performing definitive repair. An adequately
0002141491.INDD 369 7/14/2014 8:15:34 PM

controlled bile leak with total bile diversion is well tolerated for several
weeks; however, both electrolytes and volume status must be monitored
and optimized. The temptation to reexplore patients early with a major BDI
associated with a bile leak should be avoided as the associated inflamma-
tion makes dissection and anastomosis quite difficult. Usually 4 to 6 weeks
is a suitable time to plan definitive reconstruction. In those patients with a
major bile duct transection presenting with obstruction without a bile leak,
repair can be completed early after defining the anatomy.
OUTCOMES/FOLLOW-UP
Management of both postoperative bile leaks and bile duct injuries requires
a multidisciplinary team. Minor bile leaks that are appropriately man-
aged have excellent outcomes with nonoperative success rates over 90%.
Bile duct injuries even when treated at major centers still are associated
with significant morbidity and even mortality. Delay in diagnosis, failure to
control the bile leak, and uncontrolled sepsis can lead to a fatal outcome
particularly in elderly patients. Multiple series reporting outcomes follow-
ing repair have reported postoperative morbidity rates from 25% to 40%
and mortality rates of less than 2%. Complications include postoperative
wound infections, cholangitis, septicemia, intra-abdominal abscess, anas-
tomotic leak, biliary fistula, and hepatic insufficiency. In most patients, IR
plays an important role in managing complications without the need for
reoperation.
Fortunately, the long-term results for BDI following biliary reconstruc-
tion at major referral centers are generally excellent with successful out-
comes reported in 85% to 95% of cases with follow-up in excess of 5 years.
The development of liver failure or the need for liver transplantation is very
rare for isolated bile duct injuries (without associated vascular injuries) if
timely and appropriate treatment is provided. Finally, for most patients,
quality of life following BDI with successful repair eventually returns to lev-
els of healthy controls, which is particularly important as the majority of
patients in most series are middle-aged women.
CONCLUSION
Biliary leakage is a complication that usually occurs after interventions
on the gallbladder, bile ducts, and liver as well as after abdominal trauma.
The overwhelming majority of biliary leaks are minor and will resolve with
minimal intervention and morbidity. Advances in imaging and minimally
invasive techniques permit nonoperative treatment in most cases. A BDI is
a major complication with significant deleterious effects to the patient with
respect to morbidity, cost of medical care, and disruption of short-term
quality of life. Prompt recognition, appropriate management, and definitive
repair are necessary to ensure the best possible outcome.
Suggested Readings
Branum G, Schmitt C, Baillie J, et al. Management of major biliary complications after
laparoscopic cholecystectomy. Ann Surg 1993;217:532.
Flum DR, Cheadle A, Prela C, et al. Bile duct injury during cholecystectomy and sur-
vival in Medicare beneficiaries. JAMA 2003;290:2168–2173.
Flum DR, Dellinger EP, Dheadle A, et al. Intraoperative cholangiography and risk of
common bile duct injury during cholecystectomy. JAMA 2003;289:1639–1644.
Lillemoe KD, Melton GB, Cameron JL, et al. Postoperative bile duct strictures: man-
agement and outcome in the 1990s. Ann Surg 2000;232:430.
Massarweh NN, Flum DR. Role of intraoperative cholangiography in avoiding bile
duct injury. J Am Coll Surg 2007;204:656.
002141491.INDD 370 7/14/2014 8:15:34 PM

Melton GB, Lillemoe KD, Cameron JL, et al. Major bile duct injuries associated with
laparoscopic cholecystectomy: effect of surgical repair on quality of life. Ann Surg
2002;235:888.
Pitt HA, Sherman S, Johnson MS, et al. Improved outcomes of bile duct injuries in the
21st century. Ann Surg 2013;258:490.
Sheffield KM, Riall TS, Han Y, et al. Association between cholecystectomy with vs
without intraoperative cholangiography and risk of common duct injury. JAMA
2013;310:812–820.
Sicklick JK, Camp MS, Lillemoe KD, et al. Surgical management of bile duct injuries
sustained during laparoscopic cholecystectomy: perioperative results in 200
patients. Ann Surg 2005;241:786.
Stewart L, Way LW. Bile duct injuries during laparoscopic cholecystectomy: factors
that influence the results of treatment. Arch Surg 1995;130:1123–1129.
Strasberg SM, Hertl M, Soper NJ. An analysis of the problem of biliary injury during
laparoscopic cholecystectomy. J Am Coll Surg 1995;180:101.
Strasberg SM. Biliary injury in laparoscopic surgery: Part 2. Changing the culture of
cholecystectomy. J Am Coll Surg 2005;201:604.
Walsh RM, Henderson JM, Vogt DP, et al. Long-term outcome of biliary reconstruction
for bile duct injuries from laparoscopic cholecystectomies. Surgery 2007;142:450.
Way LW, Stewart L, Gantert W, et al. Causes and prevention of laparoscopic bile duct
injuries. Ann Surg 2003;237:460–469.
Wohl W, Brandt MM, Hammila M, et al. Diagnosis and management of bile leaks after
blunt liver injury. Surgery 2005;138:742–748.
0002141491.INDD 371 7/14/2014 8:15:34 PM

SECTION IV
Multidisciplinary
HPB Care
002086390.INDD 372 7/15/2014 11:21:33 AM

30 Liver and Pancreas
Transplantation
Richard S. Mangus and Jonathan A. Fridell
BRIEF HISTORY OF LIVER TRANSPLANTATION

Though orthotopic liver transplantation (LT) first became a reality in the
early 1960s, outcomes were inadequate to make it the standard of care
for end-stage liver disease until the 1980s. Two primary factors drove the
ultimate success of this procedure: (1) the introduction of calcineurin
inhibitors (cyclosporine) to provide adequate, stable, long-term immu-
nosuppression and (2) the passage of brain death laws that drastically
improved the quality of deceased donor organs. Subsequent improvements
in organ preservation, surgical technique, and intensive care unit manage-
ment have now led to survival approaching 90% at 1 year and 60% to 70% at
10 years. This success has resulted in an unprecedented demand for LT and
has driven the field of LT in search of more organs. The primary expansion
of donors for LT has come from (1) the use of nonideal (extended criteria)
deceased donors and (2) the use of partial grafts from both deceased and
living donors.
INDICATIONS AND CONTRAINDICATIONS FOR LIVER

TRANSPLANTATION
The primary indications for LT are separated into two disease mecha-
nisms: cholestatic and noncholestatic liver disease. Additionally, LT may
be indicated for certain tumors that meet specific criteria (Table 30.1).
Simultaneous transplantation of the liver with other solid organs has
been well described, including the kidney, pancreas, lung, heart, and
intestine. Unlike kidney and pancreas grafts that are allocated accord-
ing to time on the transplant wait list, liver allografts are allocated solely
based upon severity of illness. The model for end-stage liver disease score
(MELD) was developed to predict the risk of death while awaiting LT. The
MELD score is calculated from three commonly measured laboratory
values (international normalized ratio [INR], serum total bilirubin, and
serum creatinine) and serves to assess synthetic (INR) and excretory (bili-
rubin) function of the liver as well as systemic decompensation (creati-
nine). The MELD score ranges from 6 (low) to 40 (maximum), and patients
are generally not transplanted until their score is greater than 15. MELD
exception points may be granted for the presence of malignant tumor
(hepatocellular carcinoma) or quality of life issues.
Most contraindications to LT are relative but may include active sub-
stance abuse, systemic infection, cancer, HIV infection, poorly managed
psychiatric disorders, lack of social support, and significant disease of the
cardiopulmonary or neurologic systems. Surveillance for hepatocellu-
lar carcinoma (HCC) is mandatory for all patients with cirrhosis; 70% to
90% of all HCC cases arise in the cirrhotic liver. Scoring systems have been
developed to determine acceptable HCC tumor volume that still permits
373
0002086390.INDD 373 7/15/2014 11:21:33 AM

374 Section IV / Multidisciplinary HPB Care
TABLE
Common Diseases for Which Liver Transplantation May Be
30.1 Indicated
Cholestatic Liver Noncholestatic Liver

Disease Disease Other
Primary sclerosing Viral hepatitis Hepatocellular
cholangitis (PSC) carcinoma
Primary and secondary Alcoholic liver disease (ALD) Cholangiocarcinoma
biliary cirrhosis
(PBC)
Biliary atresia Nonalcoholic fatty liver Acute liver failure
disease (NAFLD)
Cystic fibrosis Autoimmune hepatitis
Hepatic toxicity from chronic
drug or toxin exposure
Metabolic or genetic
disorders
LT, the two most common of which are the Milan and the University of
California, San Francisco (UCSF) criteria (Table 30.2). Liver-directed thera-
pies for HCC such as transarterial embolization and radiofrequency abla-
tion are useful and serve to control disease burden while awaiting LT; these
therapies may also be used to downstage HCC volume to meet transplant
criteria.
End-stage liver disease is frequently accompanied by a host of comor-
bidities that may impact the outcome of the transplant. Hepatorenal syn-
drome (HRS) is a condition in which chronic liver disease is associated
with renal insufficiency. Though some of the effects of severe HRS may be
reversed with LT, patients do not generally return to completely normal
renal function. Hepatopulmonary syndrome (HPS) results from differences
in vascular pressures related to the primary liver disease. The vasodila-
tion resulting from liver failure leads to a direct vasodilatory effect in the
lungs, which results in increased blood flow in relation to ventilation and
a ventilation–perfusion mismatch. This is seen clinically as a right-to-left
shunt, and the patient experiences dyspnea. As with HRS, HPS is an impor-
tant marker of disease severity, and prognosis is poor without a timely LT.
Symptoms of HPS improve markedly with LT.
TABLE
Criteria for Liver Transplantation in Patients with Hepatocellular
30.2 Carcinoma
Milan Criteria UCSF Criteria

Number and Solitary tumor with Solitary tumor with
size of tumors maximum diameter of maximum diameter of
5 cm 6.5 cm
Up to three tumor nodules, Up to three tumor nodules,
each of which is <3 cm, with largest nodule
with no vascular <4.5 cm, or a total tumor
invasion or extrahepatic diameter <8 cm, with
metastases no vascular invasion or
extrahepatic metastases
002086390.INDD 374 7/15/2014 11:21:33 AM

Chapter 30 / Liver and Pancreas Transplantation 375
PROCEDURE/TECHNIQUE
Organ Procurement from Deceased Donors
The organ procurement operation from a deceased donor is standard-
ized with little variation among surgeons. Safe preservation of the organs
is dependent upon two primary principles: (1) rapid organ exsanguina-
tion and vascular flushing with an appropriate preservation solution and
(2) rapid organ cooling. After the organs have been prepared in the donor,
an aortic cannula is placed that is used to flush the arterial system. Some
surgeons choose to place a second cannula in the portal vein (through the
inferior mesenteric vein) to flush the portomesenteric system. The abdom-
inal aorta is clamped superior to the celiac trunk and also near the iliac
bifurcation to isolate the abdominal organs. The outflow for the blood and
preservation solution is generally through the suprahepatic vena cava at
its junction with the right atrium. As the clamps are placed and the pres-
ervation solution is infused, iced normal saline is placed throughout the
abdominal cavity to topically cool the organs. The heart arrests during this
process, and the organs are then removed. For living donor partial liver
procurement, the liver resection is performed in the donor with dissection
completed while vascular inflow and outflow remains intact. Clamps are
applied simultaneously, and the donor graft is removed, followed by rapid
flushing and cooling (Fig. 30.1A and B). Donation after cardiac death (DCD)
is currently used in 4% of deceased donors and requires complete cessation
of cardiopulmonary function prior to organ procurement. This necessar-
ily results in several minutes of graft warm ischemia time, which results
in inferior outcomes. In all cases, once procured, the liver allograft must
be transplanted and reperfused within 12 hours, after which there is an
increased risk of complications.
Liver Transplant Operation

There are three components to the liver transplant operation: (1) prepara-
tion of the transplant organ, (2) recipient hepatectomy, and (3) implanta-
tion of the graft. Preparation of the graft involves removal of residual tissue
from the donor operation including residual diaphragm and pericardium
around the vena cava and hepatic veins. The gallbladder is always removed
from the donor liver, and the cystic duct is ligated. Accessory hepatic arter-
ies are carefully dissected and preserved. An accessory right hepatic artery
requires reconstruction prior to transplantation. Frequently, an accessory
left hepatic artery cannot be safely reconstructed, because of small size,
and is dissected and preserved in situ.
The recipient hepatectomy is performed in one of two ways: (1) con-
ventional (standard bicaval) or (2) the piggyback or “cava-preserving” tech-
nique (Fig. 30.2A and B). Both approaches to the hepatectomy require initial
takedown of the falciform and gastrohepatic ligaments, followed by dissec-
tion of the hilum of the liver with transaction of the hepatic artery, common
bile duct, and portal vein. The conventional approach then proceeds with
clamping of the vena cava above and below the liver with transaction of
the vena cava between the clamps and removal of the liver. This technique
necessarily requires complete clamping of the vena cava for a prolonged
period of time. The piggyback technique varies from this approach with
no clamping of the vena cava required. For the piggyback hepatectomy,
the liver is carefully retracted away from the vena cava with perforating
branches between the vena cava and the liver individually ligated and tran-
sected. Eventually, the liver remains attached only by the hepatic veins. The
veins are clamped and transected, and the liver is removed. The piggyback
technique is technically more difficult than the conventional approach, but
0002086390.INDD 375 7/15/2014 11:21:34 AM

FIGURE 30.1 A. Living donor left lateral segment graft for a pediatric recipient. B. Living
donor right lobe graft for an adult recipient.
002086390.INDD 376 7/15/2014 11:21:35 AM

FIGURE 30.2 A. Conventional or standard bicaval anastomosis of the vena cava.

B. Piggyback or “cava-preserving” technique for LT.
0002086390.INDD 377 7/15/2014 11:21:36 AM

the recipient tends to remain more hemodynamically stable as the preload

to the heart from the lower body is never interrupted. Some surgeons who
use the piggyback approach construct a temporary portacaval shunt to
decompress the portomesenteric system until the time of liver allograft
reperfusion. The piggyback approach may be the preferred method in
high-risk patients such as the elderly, those with poor physiologic reserve,
or patients who are hemodynamically unstable. The piggyback technique
tends to preserve hemodynamic and physiologic stability throughout the
transplant, which may then be associated with less perioperative morbidity
and mortality.
Finally, the transplant is performed. For the conventional approach,
the vena cava must be reanastomosed both above and below the liver. For
the piggyback technique, the liver outflow is through the clamped hepatic
veins so that a single anastomosis is required. This reduction from two to
one vena cava anastomosis can decrease critical warm ischemia time by as
much as 5 to 10 minutes. Next, the portal vein and hepatic artery are anas-
tomosed. Finally, the common bile duct can be anastomosed, after the liver
has been fully reperfused, either to the recipient common bile duct or to a
Roux-en-Y limb of the small intestine. At our center, 90% of our transplants
employ a primary duct-to-duct reconstruction, and we never use a T tube.
In countries with limited access to deceased donors, partial liver
allografts from living donors may be the only option available for patients in
need of transplantation. Also, splitting of a whole deceased donor liver can
result in two viable grafts, though each split portion carries an increased
risk of complications when compared to a whole organ allograft. The living
donor partial liver transplant requires use of the piggyback hepatectomy
because there is no vena cava available from the living donor. Therefore,
the donor hepatic vein is anastomosed to the recipient hepatic veins, with
portal and arterial inflow constructed directly to the native portal vein and
hepatic artery (Fig. 30.3). The bile duct is transected close to the liver graft
and is generally reconstructed with a Roux-en-Y limb of jejunum.
COMPLICATIONS
Immediate posttransplant complications may result from technical issues
related to the transplant procedure, poor function of the graft, or recipient
health issues. The most critical early posttransplant complication is throm-
bosis of the hepatic artery or the portal vein. Thrombosis of the hepatic vein
anastomosis is very uncommon. The liver requires both hepatic artery and
portal vein inflow for survival, and thrombosis of either anastomosis places
the graft at imminent risk of failure. Doppler ultrasound is utilized post-LT
to assure that vascular flow remains adequate. A loss of flow in either vessel
generally necessitates emergent return to the operating room to reestab-
lish flow. Posttransplant bleeding in the first 24 to 48 hours is not uncom-
mon and requires reexploration to identify and control the source. Primary
nonfunction (PNF) is a definition of exclusion in which there is graft failure
within 7 days of the transplant with no other identifiable cause. PNF occurs
in 1% to 5% of liver transplants and may be related to prolonged ischemia
time, old donor age, and severe steatosis. Small for size graft is a syndrome
seen almost exclusively in partial LT. The small size of the transplant graft in
relation to the recipient blood flow results in persistent portal hypertension
and liver congestion as the portomesenteric flow is too great for the partial
liver graft segment. This syndrome is a major cause of early graft loss in liv-
ing donor transplantation.
Biliary complications may be seen within the first week posttrans-
plant but may also occur several weeks or months later. There are three
002086390.INDD 378 7/15/2014 11:21:36 AM

FIGURE 30.3 Implantation of the right lobe graft into an adult.
primary complications of the bile duct: (1) biliary leak, (2) anastomotic
stricture, and (3) intrahepatic strictures. Biliary leak is encountered in the
perioperative period and may be technical in nature or may result from bile
duct necrosis. Biliary complications are particularly problematic in living
donor LT and are a major cause of graft loss and patient death in this popu-
lation. Most anastomotic strictures, and many anastomotic leaks, can be
treated nonoperatively with endoscopic retrograde cholangiopancreatog-
raphy or percutaneous transhepatic cholangiography. Through these tech-
niques, stents can be placed across the anastomosis to facilitate drainage,
and these can be increased in size over time to dilate the duct. Intrahepatic
strictures appear as multiple diffuse strictures within a region of the liver or
across the entire liver. Diffuse intrahepatic structuring points to a systemic
problem usually related to arterial blood flow. Liver allografts procured
from deceased donors using a DCD protocol are at increased risk of diffuse
intrahepatic strictures. Patients who develop diffuse intrahepatic biliary
strictures have a high rate of graft loss and frequently require retransplan-
tation within 1 year.
An important impediment to success in the field of LT is the side
effects of immunosuppressive drugs. These powerful agents, though effec-
tive at preventing rejection, continue to have major side effects that impact
on long-term patient morbidity and mortality. Many centers now use
0002086390.INDD 379 7/15/2014 11:21:37 AM

a ntibody-based immunosuppression induction at the time of transplant,

followed by one, two, or three drug maintenance regimens. Rejection of the
liver allograft is less common than that seen for other solid organs. In fact,
the liver appears to lessen the rejection risk of other organs when they are
transplanted simultaneously. The reason for this finding is unclear. Because
the liver transplant procedure itself is clean, intraabdominal infections
post-LT are rare. Bacterial wound infections, however, are common for a
variety of reasons. The liver transplant incision is quite large, and patients
with liver failure are malnourished and have poor healing capacity. Many
liver transplant recipients receive perioperative steroids, which impacts
wound healing directly. Finally, clinical comorbidities such as diabetes and
obesity contribute to poor wound healing and infection. The most common
opportunistic viral infection in the liver transplant patient is cytomegalovi-
rus (CMV) infection with a 1-year risk of 5% to 20%. Anti-CMV prophylaxis
with valganciclovir and minimization of immune suppression lower the
rate of this infection. Fungal infections can occur in up to 20% of patients
and are most commonly seen in the mouth, esophagus, and urine. These
infections almost always result from Candida species and respond to stan-
dard treatments.
Nearly all post-LT immunosuppression protocols utilize calcineurin
inhibitors as the primary immunosuppressant drug. Use of these agents
has resulted in nephrotoxicity being one of the primary long-term major
complications associated with solid organ transplantation. Finally, the
immune system plays a critical role in neoplasm surveillance in the human
body. The chronic immune suppression required by all transplant patients
places them at increased risk for the development of cancer, and this is well
documented. The cancers found in transplant patients tend to mirror those
seen in the general population, but with increased frequency. Therefore, the
cancer most commonly seen in transplant patients is skin cancer (squa-
mous and basal cell). Patients with HCC at the time of transplant have a
risk of HCC recurrence posttransplant, but their survival is similar to that
for non-HCC patients if they are within the Milan criteria at the time of
transplant. Posttransplant lymphoproliferative disorder (PTLD) is well
described and appears to have increasing risk with increasing levels of
immune suppression.
PANCREAS TRANSPLANTATION
Introduction
Pancreas transplantation can restore normal insulin secretion and eugly-
cemia in insulinopenic diabetic recipients. Since the first pancreas trans-
plant performed by Lillehei and Kelly at the University of Minnesota in
1966, pancreas transplantation has evolved to the point where there are
more than 1,500 procedures performed annually in the United States and it
is routinely offered at most abdominal organ transplant centers. Outcomes
have improved steadily over the decades, owing to improvements in organ
preservation, surgical technique, postoperative care, and immunosuppres-
sion. Nonetheless, the need for a major operation and the requirement
for lifelong immunosuppression have limited the application of pancreas
transplantation to very select diabetic populations.
INDICATIONS
The majority of pancreas transplants are performed in association with
another organ transplant in recipients with type 1 diabetes mellitus. The
other organ transplanted is almost always a kidney transplant for end-stage
002086390.INDD 380 7/15/2014 11:21:37 AM

diabetic nephropathy. These patients are already committed to lifelong

immunosuppression for the renal allograft and therefore only have the added
surgical risk of implanting the pancreas. Simultaneous liver–pancreas and
lung–pancreas transplants have also been reported applying the same phi-
losophy. The kidney and pancreas transplants can be performed either simul-
taneously (simultaneous pancreas and kidney [SPK]) from a single cadaveric
donor or sequentially if a potential living donor can be identified for the kid-
ney allograft (pancreas after kidney [PAK]). A few select centers offer simul-
taneous cadaveric pancreas and living donor kidney transplantation (SPLK).
Less frequently, a pancreas transplant alone (PTA) is offered for diffi-
cult to control and potentially life-threatening diabetes mellitus, usually with
documented episodes of hypoglycemia unawareness. In these instances, the
constant risk of untreated hypoglycemia that can result in hypoxic brain
injury, accidents, and mortality outweighs the risk of the procedure and the
immunosuppression. PTA is also offered for recipients who had previously
undergone total pancreatectomy and exhibit poor glycemic control. In these
cases, enteric exocrine drainage has the additional advantage in this recipi-
ent population of restoring exocrine function and improving nutrition.
CONTRAINDICATIONS
Absolute contraindications to pancreas transplantation include substance
or alcohol abuse, uncorrectable cardiac disease, and active malignancy or
infection. Prior to listing, all patients undergo cardiac evaluation, including
a cardiac stress test with coronary angiography investigation of all abnormal
studies. Only candidates with normal or corrected cardiac status are consid-
ered eligible. There is no specific upper limit for age or body mass index that
contraindicates pancreas transplantation, although these are both features
that may increase the risk of the procedure. Similarly, although not consid-
ered a contraindication, candidates who have undergone extensive prior
abdominal operations or with advanced atherosclerotic arterial disease cer-
tainly present technical challenges that increase the risk of surgery.
SURGICAL TECHNIQUE
Pancreas allografts are procured following an aortic flush with preservation
solution. The spleen is usually retained with the pancreas until implantation
to avoid direct handling of the pancreas. The duodenum is also included
with the allograft in order to facilitate creation of exocrine drainage. The
backbench preparation of the pancreas includes a donor splenectomy and
reconstruction of the donor superior mesenteric and splenic arteries with
a donor iliac artery Y graft. Ultimately, when ready for implantation, the
iliac artery Y graft will serve as the arterial inflow and the portal vein the
venous outflow.
Pancreas transplantation is usually performed through a midline
incision. SPK transplants are routinely performed with the kidney on the
left and the pancreas on the right, although ipsilateral placement with both
organs on the right side has also been described. Surgical techniques for
pancreas allograft implantation have evolved with time. The original pan-
creas allograft was implanted with a ligated pancreatic duct. The subse-
quent series of transplants were performed with enteric exocrine drainage
and were plagued with early allograft failures because the surgical tech-
niques, organ preservation, and immunosuppression were still in their
infancy. The introduction of bladder drainage of the exocrine secretions
in the 1980s had several advantages over enteric drainage, particularly the
ability to monitor exocrine function by measuring urinary amylase excre-
tion. Furthermore, anastomotic leak from a bladder-drained pancreas
0002086390.INDD 381 7/15/2014 11:21:37 AM

is much easier to manage compared to an enteric leak. Unfortunately,

bladder drainage has its own separate list of complications, particularly met-
abolic complications related to the loss of bicarbonate in the urine, cystitis,
urethritis, and reflux pancreatitis in recipients with neuropathic bladders.
Recently, as graft survival outcomes following pancreas transplantation
have improved, enteric drainage has become the exocrine drainage proce-
dure of choice at most centers. In fact, it has also become fairly common
for the pancreas to be implanted with the head oriented toward the upper
abdomen in order to facilitate a tension-free enteric anastomosis (Fig. 30.4).
Endocrine graft drainage through the portal vein is most commonly
performed to the systemic venous circulation, usually to the right external
or common iliac veins or vena cava. An alternative technique is mesenteric
drainage through the recipient superior mesenteric vein. Regardless of the
venous outflow, the donor iliac artery Y graft is still most commonly anasto-
mosed to the right external or common iliac artery or, less often, the aorta.
The theoretical advantage of mesenteric drainage is that it more closely
imitates the normal endocrine physiology as the insulin is drained directly
into the portal vein, as it is in the native setting. This allows for the first-pass
FIGURE 30.4 Ipsilateral placement of simultaneously transplanted pan-

creas and kidney allografts. The iliac artery extension Y graft and portal
vein from the pancreas are anastomosed to the recipient common iliac
artery and vein, respectively.
002086390.INDD 382 7/15/2014 11:21:37 AM

effect through the liver, where approximately half of the insulin is removed
resulting in relatively lower systemic insulin levels. Both endocrine drain-
age techniques establish normoglycemia, although with different systemic
insulin levels, and similar graft and patient survival.
Immunosuppression
A detailed discussion of immunosuppression is beyond the scope of this chap-
ter, but some highlights are provided. The most common protocols include
induction immunosuppression with depleting antibodies such as alemtu-
zumab (Campath), a monoclonal antibody to CD52, or rabbit antithymocyte
globulin (Thymoglobulin). Maintenance immunosuppression most com-
monly includes a combination of tacrolimus and mycophenolate mofetil or
sirolimus. It has also become a relatively common practice to avoid steroids as
maintenance immunosuppression. In addition to the toxicities of the various
immunosuppression agents, there is also an associated risk for opportunistic
infections, particularly CMV and BK virus in kidney and pancreas transplant
recipients, and opportunistic malignancies such as skin cancers and lym-
phoma (including PTLD). Prophylaxis against CMV and Pneumocystis jiroveci
is routinely required following induction immunosuppression.
COMPLICATIONS
Technical failure, defined by the International Pancreas Transplant Registry
as graft loss secondary to vascular thrombosis, bleeding, anastomotic leaks,
or infection/pancreatitis, is responsible for more than half of all pancreas
grafts lost in the first 6 months following transplantation, representing
approximately 8% to 10% of transplants. Thrombosis accounts for more than
half of these technical failures and may be influenced by donor and recipient
factors, preservation and ischemic injury, immunologic issues, and surgical
technique. Timely re-exploration is critical because, if there is indeed vascu-
lar thrombosis, the resulting allograft inflammation may initiate a systemic
response with consequent recipient hemodynamic compromise or seg-
ments of the clot may form pulmonary emboli if the portal vein is systemi-
cally drained. Additionally, if explored early enough, vascular thrombectomy
with graft salvage is occasionally possible. In cases where allograft pancre-
atectomy is required, immediate retransplantation is an option that offers
the motivated patient an opportunity to regain the benefits of a functional
allograft.
Postoperative hemorrhage can present with intraperitoneal bleeding
from the pancreas allograft or retroperitoneum or with gastrointestinal
bleeding at the site of an enteric anastomosis. If the enteric anastomosis is
created at the level of the proximal jejunum, this is approachable endoscop-
ically and can often be treated without returning to the operating room. It is
a good practice to intubate the recipient prior to push enteroscopy in order
to protect the airway. In the event of severe gastrointestinal bleeding for
which angiography is considered, it is essential that it be communicated to
the interventional radiologist that the iliac arteries be interrogated in addi-
tion to the mesenteric vessels as the donor duodenum or an arterioenteric
fistula may be the source of bleeding. An option to manage severe bleeding
from the pancreas allograft endovascularly is placement of a covered stent
in the recipient iliac artery to totally bypass the takeoff of the donor artery
prior to emergency laparotomy and accept the loss of the pancreas allograft
in the event of life-threatening bleeding.
As with other pancreatic operations and enteric anastomoses,
enteric or pancreatic leaks may develop and usually present with an
abscess or fistula. Options for management include operative and
0002086390.INDD 383 7/15/2014 11:21:37 AM

percutaneous drainage. A drained abscess or fistula with low output

will usually heal with time, although if the fluid collects around the area
of the vascular anastomoses, it may predispose to the formation of arte-
rial pseudoaneurysm. A high index of suspicion is required to diagnose
and treat this life-threatening complication; any bleeding into an abscess
drain should be presumed to represent a herald bleed until pseudoaneurysm
is ruled out radiologically. Pseudoaneurysms can be managed endovas-
cularly, and graft salvage, although uncommon in this setting, has been
described.
Small bowel obstruction (SBO) can complicate any laparotomy. In
addition to adhesions, two etiologies of SBO are specific to pancreas trans-
plantation. First of all, the enteric anastomosis, particularly when the pan-
creas is oriented with the head upward, creates an internal hernia defect
that is prone to incarceration and intestinal volvulus if not specifically
closed at the time of transplantation (Fig. 30.5). Second, the anterior sur-
face of the pancreas in the retroperitoneum presents a raw surface that is
prone to adhesion formation. It is our own practice to cover the pancreas
with a sheet of adhesion barrier at the end of the operation. Similarly, the
FIGURE 30.5 Internal herniation of the small intestine behind the pancreas allograft
through a defect created at the time of transplantation. The opening is defined by the aorta
and iliac artery posteriorly, the small bowel mesentery superiorly, the pancreas and enteric
anastomosis anteriorly, and the pancreatic vascular anastomoses inferiorly.
002086390.INDD 384 7/15/2014 11:21:37 AM

intraperitoneal renal allograft should be covered with colon in order to

prevent contact with the small intestine.
OUTCOMES
In the United States, approximately 1,200 to 1,500 pancreas transplants
are performed annually. SPK makes up the majority of the pancreas trans-
plants performed, accounting for approximately two-thirds of the total.
Patient survival is currently similar for all three categories of pancreas
transplantation, approximately 96% at 1 year (Fig. 30.6). Pancreas allograft
survival has historically been superior for SPK compared to isolated pan-
creas transplantation (85% vs. 78% at 1 year) (Fig. 30.7). The difference in
graft survival is often attributed to increased immunologic pancreas graft
loss in isolated pancreas allograft recipients (PAK and PTA) compared to
SPK recipients (5% vs. 2% at 1 year). This increased pancreas graft loss to
rejection has been attributed to delayed detection and initiation of treat-
ment for rejection because these recipients lack the early warning of a
deterioration in renal function, which may be considered a harbinger of
pancreas rejection in SPK recipients. Recent data from large centers using
depleting antibody induction and steroid-free maintenance immuno-
suppression protocols suggest that we have entered a new era of immu-
nosuppression where immunologic graft loss is becoming an infrequent
occurrence. With this cause of graft failure practically eliminated, it is
likely that pancreas allograft survival in isolated pancreas transplantation
will begin to approach that of SPK. In fact, in the most recent update of
the International Pancreas Transplant Registry, the 1-year graft survival for
all three categories of pancreas transplantation was similar at 89% to 90%
for programs using anti–T-cell antibody induction and maintenance with
tacrolimus and sirolimus.
Patient survival
USA primary DD pancreas transplants 1/1/2006 – 12/31/2010

%
100
90
80
Category n 1Yr Surv.

70 PTA 433 96.4%
PAK 800 96.0%
SPK 4,129 95.8%
60
0 6 12 18 24 30 36
Months Posttransplant
3/11
IPTR/UNOS
FIGURE 30.6 Patient survival following primary deceased donor pancreas transplanta-
tion in the United States between 1/1/2006 and 12/31/2010. Pancreas transplants are
categorized by type as pancreas transplant alone (PTA), pancreas after kidney (PAK), and
simultaneous pancreas and kidney (SPK). (Reprinted with permission from the International
Pancreas Transplant Registry.)
0002086390.INDD 385 7/15/2014 11:21:38 AM

Pancreas graft function
USA DD primary pancreas transplants 1/1/2005 – 12/31/2010
%
100
80
60
PAK
PTA
40
SPK
Cat. n 1Yr Fxn
SPK 4,122 85.5%
20 PAK 798 79.9%
PTA 433 77.8% p £ 0.0001
0
0 6 12 18 24 30 36
Months posttransplant
3/11
IPTR/UNOS
FIGURE 30.7 Pancreas allograft survival following primary deceased donor pancreas
transplantation in the United States between 1/1/2005 and 12/31/2010. Pancreas trans-
plants are categorized by type as simultaneous pancreas and kidney (SPK), pancreas after
kidney (PAK), and pancreas transplant alone (PTA). (Reprinted with permission from the
International Pancreas Transplant Registry.)
Suggested Readings
Liver Transplantation
Abu-Elmagd K, Fung J, Bueno J, et al. Logistics and technique for procurement of intestinal,
pancreatic, and hepatic grafts from the same donor. Ann Surg 2000;232(5):680–687.
Fishman JA. Infection in solid-organ transplant recipients. N Engl J Med 2007;
357(25):2601–2614.
Hong JC, Yersiz H, Farmer DG, et al. Longterm outcomes for whole and segmental
liver grafts in adult and pediatric liver transplant recipients: a 10-year comparative
analysis of 2,988 cases. J Am Coll Surg 2009;208(5):682–689; discussion 689–691.
Mangus RS, Fridell JA, Vianna RM, et al. Immunosuppression induction with rab-
bit anti-thymocyte globulin with or without rituximab in 1,000 liver transplant
patients with long-term follow-up. Liver Transpl 2012;18:786–795.
Tector AJ, Mangus RS, Chestovich P, et al. Use of extended criteria livers decreases
wait time for liver transplantation without adversely impacting posttransplant
survival. Ann Surg 2006;244(3):439–450.
Pancreas Transplantation
Fridell JA, Johnson MS, Goggins WC, et al. Vascular catastrophes following pancreas
transplantation: an evolution in strategy at a single center. Clin Transplant 2011;
DOI: 10.1111/j.1399-0012.2011.01560.x.
Fridell JA, Mangus RS, Hollinger EF, et al. The case for pancreas after kidney transplan-
tation. Clin Transplant 2009;23(4):447–453.
Fridell JA, Mangus RS, Mull AB, et al. Early reexploration for suspected thrombosis
after pancreas transplantation. Clin Transplant 2011;91:902–907.
Fridell JA, Powelson JA, Sanders CE, et al. Preparation of the pancreas allograft for
transplantation. Clin Transplant 2011;25(2):E103–E112.
Fridell JA, Rogers J, Stratta RJ. The pancreas allograft donor: current status, controver-
sies, and challenges for the future. Clin Transplant 2010;24(4):433–449.
Gruessner AC, Sutherland DE, Gruessner RW. Pancreas transplantation in the United
States: a review. Curr Opin Organ Transplant 2010;15(1):93–101.
Humar A, Kandaswamy R, Granger D, et al. Decreased surgical risks of pancreas trans-
plantation in the modern era. Ann Surg 2000;231(2):269–275.
002086390.INDD 386 7/15/2014 11:21:39 AM

31 HPB Imaging (Including IR,
Ultrasound, MRI, CT, Nuclear
Medicine)
Temel Tirkes and Kumaresan Sandrasegaran
Radiology offers many imaging modalities for evaluation of diseases in the

abdomen; some include ultrasound, computerized tomography (CT), mag-
netic resonance imaging (MRI), positron emission tomography (PET) scan,
and nuclear medicine. The choice of appropriate imaging modality should
be made on an individual case basis depending on the lesion under ques-
tion, availability of imaging facilities, and cost issues. Choosing the most
appropriate imaging modality for evaluating the lesion under question
depends on the information required from imaging for clinical manage-
ment decisions. For example, while ultrasound might be an appropriate
examination for a patient with no prior medical history and increased
alkaline phosphatase and bilirubin levels, ultrasound is not an adequate
screening test for hepatocellular carcinoma in a patient with cirrhosis and
increased alpha-fetoprotein. Similarly, while multidetector CT (MDCT)
serves as the first-line imaging modality for evaluation of liver, MRI offers
an attractive radiation-free option with superior soft tissue contrast differ-
entiation. The clinical applications and benefit of MRI are rapidly expanding
due to new sequences and contrast agents. However, in most centers, MRI is
still reserved as secondary imaging modality, with such indications as char-
acterization of incidental liver lesions detected on CT or for patients with
iodinated contrast allergy.
LIVER
Ultrasound
Sonography is a safe, noninvasive, quick, and inexpensive means of eval-
uating the liver. It can be performed at the bedside, needing little patient
cooperation. The liver is ideally examined following a 6-hour fast so that
gallbladder is not contracted. The normal tissue texture of the liver is
homogeneous with fine echoes that appear as moderately short dots or
lines. Abnormal parenchymal echogenicity is judged by comparing it to
the right renal cortex. Increased echogenicity of the liver usually indicates
hepatosteatosis. Ultrasound has good sensitivity for detection of focal liver
lesions; however, specificity is not very high.
Duplex and color flow Doppler imaging improve the diagnostic capa-
bilities of ultrasound by enabling the evaluation of complex circulatory
dynamics of liver. Thrombosis, reverse flow, aneurysms, and fistulas are
demonstrated with duplex and color flow Doppler. But color Doppler stud-
ies are usually unable to evaluate the vascularity of the liver neoplasms due
to low intensity of the signals.
In liver transplant recipients, survival of the allograft depends on
patency of the hepatic artery. Changes in the normal hepatic artery wave-
form may suggest stenosis or thrombosis in these patients. Liver transplant
patients require very close follow-up for vessel patency during the immediate
postoperative period.
387
0002095524.INDD 387 7/14/2014 8:26:44 PM

Intraoperative Ultrasound
Acoustic attenuation by subcutaneous fat somewhat degrades the images
of the transabdominal ultrasound. During intraoperative ultrasound, a
high-frequency probe is directly placed on the liver surface, providing high-
resolution images. The most common application of intraoperative ultra-
sound is during surgery in patients undergoing segmental resection for
hepatic metastases. Intraoperative ultrasound can detect even the minute
liver lesions and modify the surgical management of patients. Intraoperative
ultrasound has been more widely utilized across the spectrum of HPB
surgery (see Chapter 12).
Computed Tomography
Advances in MDCT enabled very fast scan times and improved resolution
due to isotropic image acquisition. The goal of contrast enhancement is to
improve lesion visibility by increasing the relative attenuation difference
between the lesion and normal hepatic parenchyma. A routine contrast-
enhanced CT of the liver is performed during the portal phase and is usu-
ally sufficient for detection and follow-up of most hypovascular metastatic
liver lesions. Multiphasic CT is performed in the arterial, portal venous,
and delayed phases. During arterial phase, hypervascular lesions including
hepatocellular carcinoma and metastases from renal, breast, carcinoid, and
pancreatic islet cell tumors enhance earlier than the normal liver paren-
chyma. These lesions may become isodense to liver parenchyma during
portal venous or delayed phase. Some hypervascular lesions show lower
density during portal or delayed phase called “washout” phenomenon.
Detection of a hypervascular lesion that shows washout is commonly seen
in cirrhotic patients with hepatocellular carcinoma. Delayed scanning has
been employed to improve detection of intrahepatic cholangiocarcinoma
and hepatocellular carcinoma.
Noncontrast CT scan of the liver is inferior to contrast-enhanced stud-
ies for lesion detection and is not routinely performed except in certain
specific situations. Patients with liver disorders that diffusely alter hepatic
attenuation, such as hepatosteatosis, hemochromatosis, and amiodarone
toxicity should be evaluated with noncontrast CT. Noncontrast liver CT
may be indicated for evaluation of calcified lesions calcification (such as
metastases from mucinous colon carcinoma), or hemorrhage (in lesions
like hepatocellular adenomas).

The recent development of rapid acquisition techniques with excellent
image quality and tissue-specific contrast agents has made MRI the most
accurate imaging modality for the evaluation of liver disease (Fig. 31.1).
Traditionally T1- and T2-weighted sequences were performed with sup-
pression of the body fat; however, many novel pulse sequences are being
developed to decrease scan time and increase signal-to-noise ratio. MRI is
often superior to CT or ultrasound in lesion characterization but usually
reserved as a second-line modality secondary to cost.
Magnetic resonance contrast agents are crucial in liver lesion detec-
tion and characterization. There are many MR contrast agents with different
properties. Examples of extracellular contrast agents include the gadolinium
chelates, gadopentetate dimeglumine (Gd-DTPA) and gadoterate meglu-
mine (Gd-DOTA). These agents function similar to iodinated CT contrast
agents by rapidly diffusing from the intravascular space to the extracellular
space. Hepatobiliary specific agents such as gadoxatate (Gd-EOB-DTPA)
and gadobenate dimeglumine (Gd-BOPTA) increase the signal intensity of
the normal liver and hepatocyte-containing lesions such as focal nodular
002095524.INDD 388 7/14/2014 8:26:44 PM

Chapter 31 / HPB Imaging (Including IR, Ultrasound, MRI, CT, Nuclear Medicine) 389
FIGURE 31.1 Hepatocellular carcinoma visualized by multiphasic contrast-enhanced MRI

of the liver. A. Arterial phase of the study shows a mass within the left lobe of the liver
enhancing earlier than the rest of the liver parenchyma (arrow). Note the prominent intra-
tumoral arterial network. B. Portal venous phase of the examination shows that the mass is
relatively hypervascular (arrow) compared to the surrounding liver parenchyma.
(Continued)
0002095524.INDD 389 7/14/2014 8:26:45 PM

FIGURE 31.1 (Continued) C. Delayed phase of the study shows that enhancement char-
acteristic of the mass has excreted the contrast earlier than the liver parenchyma and
now shows relatively less enhancement (arrow). Early contrast uptake and early con-
trast clearance is called “washout phenomenon” and is a characteristic of hypervascular
malignancies.
hyperplasia. This increased signal intensity improves the contrast difference

between the liver parenchyma and nonhepatocellular lesions (such as metas-
tases) during the delayed phase images.
Positron Emission Tomography Scan

The liver demonstrates diffusely increased fludeoxyglucose (FDG) activ-
ity physiologically and is used as a qualitative comparison point for other
foci of uptake in the body (e.g., pancreatic and gallbladder lesions). Activity
that is equal to or greater than that in the liver raises concern for patho-
logic processes. Overall, most metastatic tumors are FDG avid and readily
detectable using PET–CT. FDG-PET imaging of hepatocellular carcinoma
is somewhat limited due to the activity of glucose-6-phosphatase, which is
found in varying degrees in this tumor. Hepatocellular carcinoma is more
FDG avid than the liver in approximately 55% of cases; it is equal to or less
avid in 30% and 15% of cases, respectively. Generally, PET detects only 50%
to 70% of hepatocellular carcinomas, but is useful in detection of distant
metastases (extrahepatic disease) as well as in evaluation of recurrence. In
FDG avid hepatocellular carcinoma, PET–CT imaging is helpful for staging,
especially in the assessment of distant metastatic disease.
GALLBLADDER
Ultrasound
Real-time transabdominal sonography is the most widely used diagnostic
examination for the gallbladder since it is quick and easy to perform. The
gallbladder is visible on sonograms in virtually all fasting patients despite
002095524.INDD 390 7/14/2014 8:26:45 PM

FIGURE 31.2 Abnormal gallbladder by ultrasound. Gray-scale image of the gallbladder and
liver is shown. There is a round gallstone (GS) located within the neck of the gallbladder;
note the posterior shadowing seen as black. There is abnormal and irregular gallbladder
wall thickening (arrow). Comet tail artifacts visualized in the gallbladder wall indicates ade-
nomyomatosis of the gallbladder. This is a diagnosis specifically made by the ultrasound.
body habitus or clinical condition. The success rate of obtaining a diagnos-

tic study is very high; therefore, ultrasound should be considered as the first
line of study for primary gallbladder pathology (Fig. 31.2). The examination
can be performed at the bedside and, since no ionizing radiation is used, is
safe in pregnant and pediatric patients.
Computed Tomography
The gallbladder in fasting patients is nearly always identified on CT scans of
the upper abdomen. Although calcified gallstones are frequently visible, CT
is not used as a primary examination for detecting gallstones because of its
lower sensitivity for gallstones and higher cost compared with ultrasound.
The main indication for CT in gallbladder disease is for the diagnosis and
staging of gallbladder carcinoma and evaluation of the complications of
cholecystitis such as perforation and pericholecystic abscess.
Cholescintigraphy
Cholescintigraphy is used primarily for the diagnosis of acute cholecysti-
tis. This study is performed by injecting Technetium-99m iminodiacetic
acid (99m Tc-IDA) compounds intravenously, and the tracer is taken up by
the liver and rapidly excreted into bile without undergoing conjugation,
allowing visualization of the gallbladder and bile ducts. Serial images are
obtained up to 1 hour. Delayed imaging up to 4 hours and possibly 24 hours
may be necessary in some instances. Nonfilling of the gallbladder on cho-
lescintigraphy indicates functional obstruction of the cystic duct and is
highly sensitive and specific for the diagnosis of acute cholecystitis in the
appropriate clinical setting (Fig. 31.3). Other uses for scintigraphic evalua-
tion of the biliary tract include assessment of gallbladder ejection fraction,
patency or bile leak after cholecystectomy, or biliary enteric anastomoses.
0002095524.INDD 391 7/14/2014 8:26:45 PM

FIGURE 31.3 Normal cholescintigram. Frontal view taken 60 minutes after injection of
radioisotope shows normal filling of the gallbladder (GB ). There is excretion of radioisotope
into the common bile duct (CBD) as well. These findings exclude possibility of acute chole-
cystitis with very high specificity. Normal uptake within the liver is seen as well.
Long-term function of Roux-en-Y biliary–enteric anastomosis may be eval-

uated by scintigraphy with delayed imaging.
BILIARY AND PANCREATIC DUCTAL SYSTEM

Ultrasound and CT, because of wide availability and ease of performance as
well as high diagnostic accuracy, are the first-line imaging techniques for
evaluation of the biliary tree. Magnetic resonance cholangiopancreatogra-
phy (MRCP) is assuming a larger role as a rapid, accurate, and noninvasive
method of evaluating the bile ducts. Currently, MRCP is replacing diag-
nostic endoscopic retrograde cholangiopancreatography (ERCP) in many
instances. Biliary scintigraphy has a limited role in biliary imaging, used
mainly for confirming acute cholecystitis and identifying bile leaks.
Ultrasound
Evaluation of the bile ducts in search of stones, masses, or obstructions is
one of the main applications for ultrasound. Real-time ultrasound read-
ily depicts dilated biliary ducts, which, in most instances, is indicative of
biliary obstruction. The extrahepatic bile duct can be seen in most patients
regardless of body habitus or clinical condition. The most distal common
bile duct (CBD) is more difficult and frequently impossible to image by
transabdominal ultrasound because of overlying gas in the duodenum and
colonic hepatic flexure.
Computed Tomography
Although ultrasound continues to be used as the initial screening test for
biliary disease, CT is as effective as is sonography in determining the caliber
002095524.INDD 392 7/14/2014 8:26:46 PM

of intra- and extrahepatic ducts. In addition, because of more complete

delineation of the total length of the CBD, CT may be more useful than sonog-
raphy for precisely defining the site and cause of biliary obstruction. Also, if
ultrasound results are equivocal, CT may be used to refine the data or con-
firm anatomy. In addition, CT can often differentiate benign from malignant
causes of obstruction and can provide guidance for biopsy and staging of
malignancies. CT cholangiography, using oral or intravenous cholecystogra-
phy agents, has been widely used in Europe and has been used on a limited
basis in some centers in the United States to evaluate ductal pathology.
Magnetic Resonance Cholangiopancreatography

MRCP is the most effective, safe, noninvasive MR imaging technique for
evaluation of the pancreaticobiliary ductal system. The usefulness and
accuracy of MRCP have been established in the evaluation of suspected
pancreaticobiliary pain, choledocholithiasis, malignant obstruction, con-
genital anomalies, and postsurgical alterations of the biliary tract and pan-
creas. Contrast administration is not necessary in the evaluation of patients
with suspected choledocholithiasis.
MRCP is mainly based on acquisition of heavily T2-weighted images,
with variants of fast spin echo (FSE) sequences; however, examination
should also include routine MRI images for a complete evaluation. The
diagnostic accuracy of MRCP is comparable to that of ERCP in the evalu-
ation of choledocholithiasis, malignant obstruction, and anatomic variants
of the pancreaticobiliary tract (Fig. 31.4). Visualization of the pancreatic
duct is often substantially improved with the use of the hormone secretin,
FIGURE 31.4 Choledocholithiasis visualized by MRCP. Coronal MRCP image shows multiple
filling defects (arrow) within the distal common bile duct causing intra- and extrahepatic
ductal dilatation.
0002095524.INDD 393 7/14/2014 8:26:46 PM

which stimulates the pancreas to secrete a significant amount of fluid while

transiently increasing sphincter of Oddi tone. Utilization of secretin is rec-
ommended in cases where detailed evaluation of the pancreatic duct is
desired. Although the role of MRCP in the evaluation of biliary diseases is
expanding, the major disadvantage of MRCP is that it is purely diagnostic
and does not provide an access for therapeutic intervention. MRCP is the
modality of choice for follow-up after some pancreatic operations such as
pancreatoduodenectomy, where ERCP is challenging.
Percutaneous Transhepatic Cholangiography and ERCP

Percutaneous transhepatic cholangiography (PTC) and ERCP are invasive
methods and usually reserved for patients requiring interventions such as
biliary drainage, stone extraction, stent placement, stricture dilatation,
and biopsy. Both techniques allow direct opacification of the biliary tree
but require invasive access and have limited diagnostic use. The choice of
procedure for direct cholangiography depends on a number of factors: the
clinical situation, the potential for therapeutic intervention, and the avail-
ability of a skilled endoscopist or interventional radiologist. Both proce-
dures allow transformation of a diagnostic study to a variety of therapeutic
maneuvers. The overall complication rate for PTC is 3.5%, which less than
the 10% complication rate seen with ERCP.
PANCREAS
The pancreas is one of the most challenging organs to image in the abdomen.
The small size of the organ and its deep location requires high-resolution
imaging. Comprehensive imaging of the pancreas should show pancreatic
and biliary ductal anatomy, help detect and characterize parenchymal dis-
ease, delineate extrapancreatic extension of a mass or inflammatory pro-
cess, and evaluate the vascular anatomy.
Ultrasound
Sonography is a practical and inexpensive option for evaluation of the pan-
creas, but it is operator dependent and usually limited in patients with large
body habitus. Ultrasound examination of the pancreas is best performed on
the fasting patient to reduce the amount of gas and food in overlying bowel.
Without patient preparation, the overlying gastric air commonly obscures
this organ, particularly the tail of the pancreas. Intraoperative sonography
is very useful.
Computed Tomography
MDCT is the first-line imaging technique for the pancreas. It is unaffected
by bowel gas or large body habitus and is widely available, fast, and rela-
tively easily performed. Multiphasic studies are the primary modality used
for evaluation and posttreatment follow-up of the pancreatic adenocar-
cinoma and hypervascular pancreatic tumors including islet cell tumors.
However, CT can be confounded by a morphologic overlap and is insensi-
tive in differentiating cystic neoplasms. In such cases, MRI of the pancreas
is recommended for further evaluation.
MRI with MRCP

MRI with MRCP examination is commonly performed together for a com-
plete evaluation of the pancreas. MRI has an advantage over CT by better
depicting the internal morphology of pancreatic cysts due to the superior
soft tissue contrast. This improved resolution facilitates recognition of
septa, nodules, and depiction of the duct. Disadvantages of MRI include
002095524.INDD 394 7/14/2014 8:26:46 PM

FIGURE 31.5 Pancreatic cystic neoplasms imaged by MRCP. This coronal maximum inten-
sity projection image shows multiple round cystic lesions arising from the pancreas (open
arrows) most likely representing an IPMN. The main pancreatic duct appears normal (short
arrows). MRI with MRCP provides superior image depiction compared to other cross-
sectional modalities such as CT.
lower spatial resolution, insensitivity to detect calcifications, and motion-

related artifacts. The contraindications to the use of MR contrast agents are
severe allergy, pregnancy, and end-stage renal dysfunction.
Cystic lesions in the pancreas constitute a diverse category including
inflammatory lesions as well as neoplasms ranging from benign lesions
to low-grade indolent neoplasia to frankly malignant tumors. One of the
most common diagnostic challenges in characterization of cystic pancre-
atic neoplasms is distinguishing isolated side branch intraductal papillary
mucinous neoplasm (IPMN) from other cystic lesions such as mucinous
cystic neoplasms or pseudocyst (Fig. 31.5). Ductal dilation may provide
evidence of chronic pancreatitis, but it may also indicate that an IPMN is
of the main duct or mixed type. These types of IPMN are more likely to be
malignant than is the more common side-branch type; therefore, a cyto-
logic evaluation may be necessary in these patients. Improvement in the
visualization of the duct has been reported with the use of the hormone
secretin, which stimulates the pancreas to secrete a significant amount of
fluid, while transiently increasing the tone of the sphincter of Oddi. MRCP
is also valuable providing noninvasive evaluation of the chronic pancre-
atitis by showing the ductal ectasia, atrophy, or strictures with upstream
dilatation (Fig. 31.6).
PET Scan
CT and MRI are used primarily to image pancreatic ductal adenocarcinoma,
but they may be limited in the setting of enlargement of the pancreatic head
without discrete mass or mass-forming pancreatitis. Metabolic imaging
may be applied to improve preoperative diagnostic accuracy and poten-
tially limit adverse outcomes from inappropriate surgical interventions.
Studies have demonstrated the relatively high sensitivity and specificity
0002095524.INDD 395 7/14/2014 8:26:46 PM

FIGURE 31.6 Pancreatic duct stricture visualized by MRCP. Coronal MRCP image demon-
strates marked dilatation of the main pancreatic the duct within the body and tail (short
arrow). This dilatation is clearly secondary to a stricture within the head of the pancreas
(long arrow).
of PET in distinguishing benign and malignant lesions in the pancreas in

comparison to CT. The primary challenge faced by PET–CT in the imag-
ing of pancreatic cancer pertains to altered glucose metabolism created by
glucose intolerance and diabetes seen in these patients. This clinical situa-
tion may create false-negative findings in patients who are hyperglycemic
or have inadequately controlled blood glucose levels. False negatives may
also result when the tumor is less than 1 cm, such as in small ampullary
carcinomas. False positives are mainly the result of inflammation secondary
to pancreatitis.
Suggested Readings
Gore RM, Levine MS. Textbook of gastrointestinal radiology, 3rd ed. Philadelphia, PA:
Saunders/Elsevier, 2008.
Semelka RC. Abdominal-pelvic MRI, 2nd ed. Hoboken, NJ: Wiley-Liss, 2006:1427, xi
Siegelman ES. Body MRI, 1st ed. Philadelphia, PA: Saunders, 2005:527, ix
Webb WR, Brant WE, Helms CA. Fundamentals of body CT, 2nd ed. Philadelphia, PA:
Saunders, 1998:363, xii.
002095524.INDD 396 7/14/2014 8:26:47 PM

32 Endoscopy in HPB Surgery
Including ERCP and EUS
Wesley D. Leung and Gregory A. Coté
INTRODUCTION
Gastrointestinal endoscopy offers a multitude of diagnostic and therapeutic
options for patients with pancreatic disease. Pancreatobiliary endoscopy
derives from endoscopic retrograde cholangiopancreatography (ERCP) and
endoscopic ultrasound (EUS), although gastroduodenal stenting is increas-
ingly utilized for patients with unresectable pancreatic cancer and gastric
outlet obstruction as well. This chapter reviews the indications, benefits,
and limitations of ERCP, EUS, and gastroduodenal stenting, with a particular
focus on pancreatic cancer and chronic pancreatitis (CP).
ENDOSCOPIC RETROGRADE CHOLANGIOPANCREATOGRAPHY

ERCP is an endoscopic technique that emerged in the 1970s where a side-
viewing upper endoscope (a.k.a., duodenoscope) is advanced per oral to
the level of the major and minor papilla, allowing instruments to access
the bile and pancreatic ducts under endoscopic and fluoroscopic guidance.
Traditionally, these ducts are opacified after the injection of iodinated contrast
and visualized fluoroscopically. In certain circumstances such as complex
choledocholithiasis or indeterminate strictures, a single-operator, fiberoptic
cholangiopancreatoscope or mother–daughter scope can be advanced into
the pancreatobiliary ducts for direct visualization. In its nascence, ERCP was
a diagnostic tool in an era where cross-sectional imaging of the pancreas and
bile ducts was extremely limited. Now, ERCP is primarily reserved for therapy,
including stone extraction, stricture dilation, or biliary drainage. ERCP has
substantial morbidity that mirrors some low- to moderate-risk surgical proce-
dures, including acute pancreatitis (5% to 15%), hemorrhage (1%), cholangitis
(1%), perforation (0.5%), and rarely death (0.1%).
ENDOSCOPIC ULTRASOUND
EUS combines a flexible endoscope with an ultrasound transducer (newer
models are electronic, although older versions are mechanical) that permits
full-thickness visualization of the gastrointestinal wall and surrounding
structures, including the pancreas and extrahepatic biliary tree. Radial echo-
endoscopes provide a circumferential view at right angles to the scope shaft,
whereas linear echoendoscopes yield a 270-degree oblique view that is in the
same plane as the scope shaft, permitting the passage of instruments under
ultrasound guidance. EUS-guided fine needle aspiration (EUS–FNA) can be
performed through a linear echoendoscope with limited risks of bleeding,
perforation, and infection, the latter occurring rarely in the setting of cyst
aspiration. Most complications from EUS and ERCP are managed medically
and are self-limiting, but serious complications such as severe necrotizing
pancreatitis, hemorrhage (refractory to endoscopic and/or percutaneous
treatment), and perforation may require surgical intervention.
397
0002086391.INDD 397 7/14/2014 8:29:36 PM

PREPROCEDURE EVALUATION
EUS and ERCP are performed most commonly on an outpatient basis.
The procedures can be performed with moderate sedation in some cases,
although facilities are increasingly utilizing monitored anesthesia care or
general anesthesia as the preferred approach. For elective cases, patients
should fast for 2 hours (clear liquids) and 8 hours (solid food) before the
procedure. Ideally, anticoagulants (e.g., warfarin) and antiplatelet agents
(e.g., clopidogrel) should be held for 5 to 7 days, particularly in the setting of
EUS–FNA or ERCP with sphincterotomy. As a general rule, an international
normalized ratio (INR) less than 1.5 and platelet count greater than 50,000
are preferred, particularly if therapeutic maneuvers are anticipated (i.e.,
endoscopic sphincterotomy or EUS–FNA). Aspirin ≤325 mg by mouth daily
probably does not impact the risk of postprocedure hemorrhage.
For ERCP, patients are preferably sedated in the prone position to opti-
mize fluoroscopic visualization and to keep the duodenoscope in a stable
“hockey-stick” configuration at the level of the papilla (Fig. 32.1). Patients
undergoing EUS are typically sedated in the left lateral position, similar
to a standard esophagogastroduodenoscopy (a.k.a., upper endoscopy).
Following either procedure, patients who are at a moderate to high risk for
FIGURE 32.1 Endoscopic retrograde cholangiopancreatography. A retrograde cholangio-

gram is performed with the duodenoscope in a “hockey-stick” configuration in the sec-
ond portion of the duodenum. This patient is a 53-year-old woman with idiopathic CP and
jaundice. She has segmental narrowing in the intrapancreatic segment of the bile duct
consistent with a CP–induced bile duct stricture.
002086391.INDD 398 7/14/2014 8:29:36 PM

Chapter 32 / Endoscopy in HPB Surgery Including ERCP and EUS 399
complications should be kept fasting or advanced to clear liquids only and

resume normal diet the next morning in the absence of concerning signs/
symptoms. Patients who are at low risk of complications can gradually
advance their diet over 4 to 6 hours.
ERCP IN THE SETTING OF PERIAMPULLARY MALIGNANCY

With improvements in cross-sectional imaging and EUS, the role of ERCP
for the diagnosis of periampullary malignancy (distal cholangiocarcinoma
and pancreatic ductal adenocarcinoma [PDAC], in particular) is increas-
ingly limited. ERCP remains the procedure of choice for biliary drainage, if
indicated, and is a medium for obtaining tissue samples.
Periampullary malignancies may present with a bile duct stricture that
can be challenging to distinguish from a benign bile duct stricture, such as
from CP. Malignant biliary strictures often have a length greater than 10 mm,
a ragged contour, fixed filling defect, and/or shouldering above the stric-
ture (abrupt transition in appearance from normal to stricture) (Fig. 32.2).
The malignant potential of proximal common hepatic duct or bifurcation
strictures is generally higher than that of distal bile duct strictures, often
due to cholangiocarcinoma or portal lymphadenopathy causing extrinsic
compression.
FIGURE 32.2 Malignant distal bile duct stricture. A 63-year-old woman presented with
painless jaundice and was found to have a pancreatic head mass on EUS performed imme-
diately prior to ERCP. EUS–FNA confirmed the presence of PDAC. Retrograde cholangiogram
reveals a discrete common bile duct stricture with abrupt “shouldering” of the upstream,
dilated common hepatic duct. This cholangiogram is highly suggestive of a malignant bile
duct stricture, as in this case.
0002086391.INDD 399 7/14/2014 8:29:37 PM

Tissue Sampling
ERCP-based tissue sampling is usually performed with a cytologic brush or
forceps biopsies. The sensitivity of cytologic brushings from bile duct stric-
tures is approximately 30% to 40% and specificity approaching 100%. Brush
cytology in primary biliary tumors has a higher sensitivity than in extrin-
sic tumors (pancreatic or metastatic) (80% vs. 35%). Fewer studies have
addressed the yield of cytologic brushings of pancreatic ducts. The accu-
racy is estimated to be 72%, and using a technique of collecting pancreatic
juice above a stricture, a sensitivity and specificity of 93% and 100%, respec-
tively, can be obtained. However, risks of pancreatitis and hemorrhage in
the upstream pancreatic duct should be considered. In patients with PSC,
diagnosing malignancy can be difficult, and cholangiography increases the
specificity and positive predictive value for cholangiocarcinoma when com-
bined with tumor markers (carbohydrate antigen 19-9) and cross-sectional
imaging. There is no clinically significant difference in diagnostic yield
between cytologic brushes produced by different manufacturers.
The sensitivity of forceps biopsy ranges from 43% to 88% with specific-
ity similar to brush cytology (approaching 100%). Flexible forceps permit
easier biliary cannulation than do stiffer ones. The combination of cytologic
brushing and forceps biopsy increased sensitivity by 15% in one report but
a smaller benefit in another study. Another study found that combination
of cytologic brushing, forceps biopsy, and endoscopic fine needle aspiration
was more sensitive (73% to 77%) than was each approach alone. Aspiration
of bile for cytologic analysis is rarely used because of low sensitivity of 6%
to 32%.
In some cases, methods such as digital image analysis (DIA) and fluo-
rescence in situ hybridization (FISH) can improve the sensitivity of cytology
alone. DIA quantifies the amount of cellular DNA by measuring the inten-
sity of nuclei stained with a dye that binds to nuclear DNA. FISH uses a fluo-
rescently labeled probe to detect chromosomal abnormalities in cells. FISH
increases sensitivity of cytology to a greater degree than does DIA. However,
there is the potential for false positives with either technique, particularly
in cases of primary sclerosing cholangitis (PSC). Therefore, a positive FISH
alone in the absence of a positive cytology specimen needs to be considered
with the rest of the clinical picture.
Miniature endoscopes can be used to visualize the bile (cholangios-
copy) and pancreatic (pancreatoscopy) ducts. Traditional mother–daughter
systems required two operators to maneuver the duodenoscope and the
cholangioscope independently. A single-operator, fiberoptic cholangioscope
(Spyglass, Boston Scientific Corp.) is available; while this facilitates direct
visualization of the bile and pancreatic ducts, its impact on confirming the
presence or absence of malignancy remains unclear.
TREATMENT OF JAUNDICE
The location of malignant biliary obstruction influences the efficacy of
endoscopic biliary drainage (distal better than perihilar). The endoscopic
management of proximal bile duct strictures (bismuth II–IV) and bile duct
obstruction are not discussed in this review. ERCP is typically preferred
to percutaneous transhepatic cholangiography (PTC) for biliary drainage
because an external drain is avoided. The short-term (<90 day) success of
ERCP in achieving biliary drainage in the setting of distal bile duct obstruc-
tion is 80% to 90%. Complications may occur in up to 10% of patients,
including cholangitis, perforation, bleeding, and post-ERCP pancreatitis.
Plastic biliary stents are relatively inexpensive and easily removable com-
pared to self-expandable metallic stents (SEMS). However, SEMS have a
002086391.INDD 400 7/14/2014 8:29:37 PM

larger diameter (8 to 10 mm as opposed to 7 to 10 Fr for plastic) and are thus

less likely to occlude prematurely (median patency rates of 6 to 9 months for
SEMS as compared to 3 months for plastic stents).
In patients with locally advanced or metastatic PDAC, the goals of
biliary drainage are palliation of symptoms and to permit systemic che-
motherapy with a normal or near-normal serum bilirubin. The treatment
of jaundice will undoubtedly relieve pruritus (if present) and may improve
other symptoms such as anorexia and indigestion. For cases of unresect-
able disease, an endoscopic approach is safer than is surgical bypass but
with reduced durability. Long-term patency of SEMS is superior to that
of plastic stents due to their larger diameter; SEMS should be placed in
patients having a life expectancy of greater than 3 months. The majority
of patients in the United States with potentially resectable periampul-
lary malignancy undergo preoperative biliary drainage despite a lack
of evidence that it reduces postoperative complications. This practice
may be explained in part by a delay in surgical consultation and concern
for higher risk of perioperative complications in the setting of jaundice.
Older studies suggest jaundice may be a predictor of postoperative infec-
tion and renal and nutritional complications. Experimental studies dem-
onstrate improved nutritional status and immune function and reduced
endotoxinemia after biliary decompression. Prospective studies evaluat-
ing the role of preoperative biliary drainage have excluded patients with
deep jaundice (serum total bilirubin ≥10 to 15 mg/dL), where the ben-
efits of drainage are expected to be the highest. Furthermore, the duration
between preoperative biliary drainage and surgery is often less than 4 weeks
despite evidence suggesting normalization of hepatocyte function after
6 weeks of decompression.
Treatment of Gastric Outlet Obstruction

Duodenal obstruction from pancreatic cancer typically is not present at
diagnosis but may develop in 15% to 20% of patients. To avoid this compli-
cation, many surgeons create a prophylactic palliative gastrojejunostomy
with a biliary bypass in those who are deemed to be unresectable at explo-
ration. Endoscopic placement of a self-expandable, gastroduodenal metal-
lic stent has lower morbidity compared to surgical gastrojejunostomy and
comparable short-term efficacy but lower long-term (>3 to 6 months) effi-
cacy due to tumor ingrowth within the interstices of the stent. In a study
comparing patients who underwent surgical gastrojejunostomy with endo-
scopic stenting, patients who underwent endoscopic palliation had similar
median survival (94 vs. 92 days) but shorter hospitalization (4 vs. 14 days),
which led to overall lower cost ($9,921 vs. $28,173).
ERCP IN THE SETTING OF CHRONIC PANCREATITIS

ERCP has evolved from a diagnostic to a therapeutic tool for patients
with CP as a result of advances in magnetic resonance cholangiopancrea-
tography (MRCP), computed tomography (CT), and EUS. Pancreatic duc-
tal changes on ERCP can be classified by the Cambridge classification as
equivocal (type I), mild to moderate (type II), and severe (type III). The
classification has been shown to correlate with the severity of exocrine
insufficiency. However, some patients with early CP have a normal pancrea-
togram. Achieving a definitive diagnosis in this population is challenging
and requires the incorporation of other imaging and pancreatic function
testing.
Pain from CP is probably multifactorial, although one suggested mech-
anism is pancreatic ductal hypertension due to pancreatic duct obstruction
0002086391.INDD 401 7/14/2014 8:29:37 PM

from a stricture or stone. Therapeutic ERCP can alleviate pain and recur-
rent pancreatitis flares with serial dilation of pancreatic duct strictures or
stone extraction. Similar to the treatment of benign bile duct strictures,
pancreatic duct strictures are managed with serial dilation and placement
of one or more plastic stents until the stricture is obliterated. This usually
requires three or more ERCPs occurring every 2 to 4 months for up to 1 year.
Pancreatic stones are more easily removed if they are small, few in number,
closer to the head of the pancreas, and not impacted. For either stricture
or stone management, a pancreatic sphincterotomy is usually required.
Stones upstream to the tail from a pancreatic duct stricture require stric-
ture dilation before extraction. Other options for stone removal include
direct pancreatoscopy, electrohydraulic lithotripsy, and extracorporeal
shock wave lithotripsy (ESWL). Notably, ESWL alone may be as effec-
tive as ESWL combined with ERCP for therapy of pancreatic duct stones.
Although studies have shown endoscopic therapy may improve pain, in CP,
the short- and long-term efficacy of endoscopic treatment are inferior to
surgical interventions aimed at relieving obstruction and achieving pain
relief.
ERCP IN THE SETTING OF ACUTE PANCREATITIS

The role of ERCP in the early management of patients with acute pancre-
atitis is limited. In patients with gallstone pancreatitis, urgent ERCP (to be
performed within 72 hours of presentation) is indicated only in the setting
of concomitant acute cholangitis or severe acute pancreatitis with biliary
obstruction. The benefits of early ERCP remain controversial. One meta-
analysis concluded early ERCP reduced complications but not mortality
in severe biliary pancreatitis and had no benefit in mild pancreatitis. Two
other meta-analyses found no benefit in morbidity or mortality from early
ERCP in biliary pancreatitis.
For all patients with suspected gallstone pancreatitis, those having a
high risk for common bile duct (CBD) stones should undergo CBD imaging
prior to discharge. High suspicion would include the presence of stones or
ductal dilation on cross-sectional imaging or persistent elevation in chole-
static labs (alkaline phosphatase, direct bilirubin). Patients at intermedi-
ate risk should undergo less invasive imaging of their extrahepatic biliary
tree in lieu of diagnostic ERCP, including EUS or MRCP. For stones less than
4 mm, EUS is more sensitive, specific, and cost-effective than is MRCP. If
EUS or MRCP confirms the presence of choledocholithiasis, ERCP should
be performed soon thereafter.
The role of ERCP with sphincter of Oddi manometry in patients with
idiopathic, recurrent acute pancreatitis requires clarification. A recent
prospective randomized clinical trial suggested no incremental benefit
of pancreatic sphincterotomy over biliary sphincterotomy alone in this
population. The benefit of empiric biliary sphincterotomy is debatable and
also requires further investigation. Individually, or in combination, MRCP
and EUS have largely replaced ERCP in diagnosing less common causes of
recurrent acute pancreatitis, including pancreas divisum, intraductal papil-
lary mucinous neoplasm, and ampullary lesions, among others.
ENDOSCOPY IN THE SETTING OF PANCREATIC FLUID COLLECTIONS

Pancreatic duct disruption and fluid collections may occur in the setting
of acute or CP. In properly selected cases, endoscopy can obviate the need
for percutaneous drains and surgery by promoting internal drainage. A
pseudocyst is an organized fluid collection, consisting of pancreatic juice
002086391.INDD 402 7/14/2014 8:29:37 PM

enclosed by a nonepithelialized tissue. Many pseudocysts can be managed

without intervention (even if over 6 cm) unless they enlarge or become
symptomatic. Traditional surgical drainage of pseudocyst carries a 10%
morbidity rate and 1% mortality; therefore, radiologic and endoscopic
drainage compete with surgery as first-line treatment. Endoscopic drain-
age can involve either transpapillary or transmural (transgastric, transduo-
denal) placement of stents. A transpapillary approach is used for smaller
pseudocysts communicating with the main pancreatic duct, whereas larger
pseudocysts or those not clearly communicating with the pancreatic duct
are drained transmurally. Pseudocyst drainage is usually performed after
allowing the cyst wall to mature for at least 4 to 6 weeks after the onset of
acute pancreatitis.
Transpapillary Drainage
Ongoing pancreatic duct disruption may heal in 78% to 92% of cases with
stent placement across the disruption. Success rates plummet to 23%
to 44% with transpapillary stents that do not bridge the leak (Fig. 32.3).
Technical success may be lower for tail disruptions and definitely in the set-
ting of complete pancreatic duct disruption.
FIGURE 32.3 Severe CP with low-grade fistula. A 45-year-old woman with severe alcohol-
induced CP presented with abdominal pain and pancreatic ascites. ERCP confirmed the
presence of a low-grade leak from the pancreatic head (arrow) with severe CP changes
in the main pancreatic duct and its side branches. The leak resolved following pancreatic
sphincterotomy and placement of a 7-Fr plastic pancreatic duct stent.
0002086391.INDD 403 7/14/2014 8:29:38 PM

Transmural Drainage
The goal of transmural drainage is to create an internal communication
between the pseudocyst and the gastric or duodenal lumen. Traditionally,
endoscopic cystogastrostomy involved piercing an endoscopically visible
bulge using a needle knife catheter and electrocautery. The risk of bleeding
is reduced from 15.7% to 4.6% if a Seldinger technique (advancing a guide-
wire through a 19-gauge needle) is used. After obtaining access to the cyst
cavity, the opening is balloon dilated, and several pigtail stents are placed.
Cyst localization by EUS facilitates drainage, particularly in the setting of
a pseudocyst that is not causing an endoscopically visible bulge; EUS has
a higher technical success and lower morbidity versus a non-EUS–guided
approach. In addition, EUS can determine if the distance between cyst and
GI tract is ideal, identify intervening blood vessels, and quantify the volume
of necrotic debris within the cyst.
Necrotic debris within a pseudocyst often necessitates débridement:
Endoscopically, this entails entering the cavity with an endoscope and
removing necrotic debris using a variety of instruments. A retrospective study
suggested that, for patients with walled-off pancreatic necrosis (WOPN),
endoscopic necrosectomy was superior to endoscopic cystogastrostomy
alone for resolution of necrosis, decreased need for adjunctive surgical or per-
cutaneous drainage, and recurrence. A step-up approach was recently com-
pared to open necrosectomy as first-line treatment for patients with proven
or suspected infected WOPN. Patients were randomized to primary open
necrosectomy and continuous postoperative lavage or to a step-up approach
(percutaneous/endoscopic drainage and if no clinical improvement repeat
drainage followed by video-assisted retroperitoneal debridement (VARD)
followed by open necrosectomy). Another recent trial compared endoscopic
necrosectomy to VARD and open surgery in the treatment of infected WOPN.
Pancreatic necrosis should ideally be addressed in the context of a multidis-
ciplinary team including pancreatic surgeons and endoscopists.
ERCP FOR THE MANAGEMENT OF POSTOPERATIVE COMPLICATIONS

The most common postoperative biliary strictures occur following liver
transplant and, less commonly, cholecystectomy. In general, the manage-
ment of benign bile duct strictures involves serial dilation using graduated
increasing size of balloons, followed by the placement of multiple plastic
stents in parallel to keep the stricture patent. For strictures involving the com-
mon hepatic duct, short-term success rates are generally greater than 90%
and have reasonable durability, with recurrence rates of approximately 10%.
Bile duct strictures occurring after liver transplant have an incidence of
4% to 16% and can be classified as anastomotic or nonanastomotic. Most anas-
tomotic strictures occur within 12 months of liver transplant. Those occurring
within 6 months of surgery respond well to endoscopic balloon dilation and
short-term (3 to 12 months) plastic stenting. Patients presenting greater than
6 months postoperatively require multiple endoscopic procedures to serially
dilate and upsize the number/maximum diameter of stents. Superior resolu-
tion rates are observed with higher total number of stents placed. In patients
with Roux-en-Y choledochojejunostomy, PTC is usually preferred to ERC since
patients require multiple procedures for serial dilation and endoscopic access
to the biliary anastomosis is challenging with Roux-en-Y anatomy.
ENDOSCOPIC ULTRASOUND
EUS for Tissue Sampling
EUS plays an important role in tissue diagnosis to confirm the presence
of pancreatic cancer and rule out metastatic lesions to the pancreas
002086391.INDD 404 7/14/2014 8:29:38 PM

(11% in one study) or surrounding structures and to diagnose nonmalignant

processes such as autoimmune or CP. In general, EUS–FNA is the most
accurate diagnostic modality with sensitivity 80% to 95% and specificity of
100% compared to CT or transabdominal ultrasound–guided FNA (sensi-
tivity 62% to 81%). In the setting of CP, sensitivity of EUS decreases to 73%.
EUS–FNA is also more accurate than is CT or magnetic resonance imaging
at detecting pancreatic masses ≤1.5 cm. The needle tract for EUS–FNA of
pancreatic head lesions lies within the surgical resection margin and thus
has a lower risk of tumor seeding. This is not the case with lesions of pan-
creatic body and tail, but to date, no data definitively show EUS–FNA nega-
tively impacts the likelihood of complete surgical resection or postoperative
recurrence. A decision to perform an EUS–FNA of a pancreatic mass should
be made in conjunction with a hepato-pancreatic-biliary (HPB) surgeon: If
the patient is a surgical candidate and the pretest probability of cancer is
greater than 90%, direct referral to surgery is also reasonable.
EUS for Staging

The accuracy of EUS for locoregional staging of pancreatic malignancy is
operator dependent. A prospective clinical trial demonstrated EUS had
no incremental value over contrast-enhanced CT scan in determining the
resectability of a pancreatic mass. A recent review of studies comparing EUS
and CT for preoperative staging of pancreatic cancer determined that it is
unclear which modality is superior for both tumor and nodal staging, though
this is less likely to impact on surgical management. A recent meta-analysis
evaluating the accuracy of individual EUS criteria for vascular invasion had
73% sensitivity and 90% specificity.
EUS for Pancreatobiliary Drainage

The technical success of deep biliary cannulation by skilled endoscopists
during ERCP is greater than 90%. The development of EUS-guided ante-
grade cholangiopancreatography refers to two different approaches to
pancreatic drainage: EUS rendezvous or direct EUS intervention. These
techniques have recently emerged as salvage methods for failed biliary
cannulation, although their utilization is only available in selected centers.
During rendezvous, EUS is used to access the bile duct transgastrically using
a 19-gauge or 22-gauge FNA needle. A guidewire is then advanced through
the needle and into the duodenum, followed by exchange of the echoen-
doscope for a duodenoscope to complete the ERCP procedure. Successful
drainage occurred in 56% of patients in a recent cohort. Direct intervention
via EUS involves creating an enterobiliary or enteropancreatic fistula. Here,
EUS is again used to direct an FNA needle into the bile or pancreatic duct
and advancing a guidewire across the obstructing lesion. Instead of switch-
ing to a duodenoscope and completing the procedure via ERCP, passage or
balloon dilation of the EUS needle track is performed followed by deploying
a stent directly. In one report, this technique had a 71% success rate but a
16% complication rate, including self-limited pneumoperitoneum and mild
to severe pancreatitis. These techniques remain experimental and should
not be performed in routine practice. However, advances in EUS-based
devices are expected to expand the role of EUS-based therapy in the future.
EUS for Palliation of Pain

The celiac plexus is one of several purported locations mediating pain
from the pancreas. Celiac plexus neurolysis (CPN) utilizes ethanol injec-
tion to ablate the celiac plexus in cases of refractory abdominal pain due to
pancreatic cancer, whereas a celiac plexus block (CPB) involves the injec-
tion of a local anesthetic with or without steroid to achieve a temporary
0002086391.INDD 405 7/14/2014 8:29:38 PM

(3 to 6 month) block in patients with CP. CPN or CPB can be achieved via
surgical and transcutaneous approaches with small but potentially sig-
nificant morbidity. EUS-guided CPN/CPB is technically feasible given the
plexus location adjacent to the celiac artery takeoff from the descending
aorta. While the celiac artery origin is easily identified during EUS, celiac
ganglia are only visualized in a minority of cases (Fig. 32.4). For patients
undergoing celiac block, a recent trial suggested the addition of triamcino-
lone did not increase the proportion of patients who reported significant
pain relief or reduce self-reported pain scores compared to injection of a
local anesthetic alone.
A recent meta-analysis suggests that EUS–CPN is safe and effective for
patients with pancreatic cancer and CP. In pancreatic cancer, nearly 80% of
patients experience pain relief after EUS–CPN; pain reduction lasts approx-
imately 20 weeks. Early consideration (i.e., during the initial diagnostic pro-
cedure in select patients) for EUS–CPN should be made in patients with
unresectable PDAC with abdominal pain requiring opioids, as it is associ-
ated with less pain and opioid requirements. In CP, the response rate and
durability are lower with EUS–CPB, with 55% of individuals having a sig-
nificant response rate initially but only 10% reporting a sustained response
after 24 weeks.
ERCP IN THE SETTING OF POSTOPERATIVE ANATOMY

ERCP for patients with postoperative anatomy (i.e., status post pancreato-
duodenectomy, Roux-en-Y gastrojejunostomy, antrectomy with Billroth I/II
reconstruction) is challenging. The optimal therapeutic approach (endo-
scopic, percutaneous, surgical, or some combination) remains unclear
but probably relates to a combination of the indication for ERCP (biliary
FIGURE 32.4 EUS-guided CPB. An EUS–FNA needle can be inserted into a celiac ganglion,
typically located anterior to the celiac artery takeoff from the aorta. Factors associated
with a better response to CPB include direct injection of celiac ganglia (when visualized)
and absence of tumor invasion of the celiac plexus. (Reproduced from Coté GA, Sherman S.
Endoscopic palliation of pancreatic cancer. Cancer J 2012;18(6):584–590.)
002086391.INDD 406 7/14/2014 8:29:38 PM

or pancreatic and the overall urgency for the procedure), specific anatomy
(length of Roux limb), and patient preference. For patients who have under-
gone Roux-en-Y gastric bypass for weight loss, endoscopy obviates the
need for percutaneous catheters but has a lower technical success rate. If
an endoscopic approach is recommended, patients with bariatric-length
Roux limbs who require pancreatic duct therapies (e.g., sphincter of Oddi
dysfunction, pancreatic duct stones, or strictures) should undergo surgical
placement of a large-bore (≥30 Fr) gastrostomy tube into the gastric rem-
nant to facilitate ERCP access. On the other hand, based on limited data,
patients requiring short-term bile duct therapy (e.g., CBD stones) may be
approached endoscopically (i.e., double balloon enteroscopy) if the antici-
pated success rate is greater than 80%.
For patients with a native papilla and short (≤30 cm) Roux limb, a
transoral approach using a duodenoscope can be attempted but is often
unsuccessful. Alternatively, a colonoscope or small bowel enteroscope may
be used to perform ERCP, but the number of ERCP accessories is limited
by the longer length of these scopes. For patients with a native papilla and
long Roux limb, surgical gastrostomy–assisted ERCP has higher success
rates than has deep enteroscopy (balloon-assisted, spiral enteroscopy) but
with higher morbidity (primarily related to complications from the surgical
gastrostomy tube). For patients without a native papilla (i.e., bilioenteric/
pancreatoenteric anastomosis) and a short Roux limb, a forward-viewing
colonoscope or enteroscope can be used successfully in most cases.
CONCLUSION
Pancreatobiliary endoscopy represents an important component of the
diagnostic and therapeutic management of patients with pancreatobi-
liary disease. Ideally, the majority of patients will be served by a multi-
disciplinary approach early in their disease course, combining the expert
opinions of abdominal radiology, medical oncology, and gastroenterology
with hepatobiliary–pancreatic surgery. A collegial working relationship
between pancreatobiliary endoscopists and their surgical colleagues is
critical since patients often require interventions from both disciplines. To
that end, surgeons should recognize the potential benefits and limitations
of pancreatobiliary endoscopy.
Suggested Readings
Anderson MA, Ben-Menachem T, Gan SI, et al. ASGE standards of practice committee:
management of antithrombotic agents for endoscopic procedures. Gastrointest
Endosc 2009;70(6):1060.
Andriulli A, Loperfido S, Napolitano G, et al. Incidence rates of post-ERCP complica-
tions: a systematic survey of prospective studies. Am J Gastroenterol 2007;102(8):
1781.
Bakker OJ, van Santvoort HC, van Brunschot S, et al; Dutch Pancreatitis Study Group.
Endoscopic transgastric vs surgical necrosectomy for infected necrotizing pan-
creatitis: a randomized trial. JAMA 2012;307(10):1053–1061.
Cahen DL, Gouma DJ, Nio Y, et al. Endoscopic versus surgical drainage of the pancre-
atic duct in chronic pancreatitis. N Engl J Med 2007;356:676–684.
Costamagna G, Tringali A, Mutignani M, et al. Endotherapy of postoperative bili-
ary strictures with multiple stents: results after more than 10 years of follow-up.
Gastrointest Endosc 2010;72(3):551–557.
Coté GA, Imperiale TF, Schmidt SE, et al. Similar efficacies of biliary, with or without
pancreatic, sphincterotomy in treatment of idiopathic recurrent acute pancreati-
tis. Gastroenterology 2012;143(6):1502–1509.
DeWitt J, Devereaux B, Chriswell M, et al. Comparison of endoscopic ultrasonography
and multidetector computed tomography for detecting and staging pancreatic
cancer. Ann Intern Med 2004;141(10):753–763.
0002086391.INDD 407 7/14/2014 8:29:38 PM

Hosono S, Ohtani H, Arimoto Y, et al. Endoscopic stenting versus surgical gastroenter-

ostomy for palliation of malignant gastroduodenal obstruction: a meta-analysis.
J Gastroenterol 2007;42(4):283–290.
Moss AC, Morris E, Mac Mathuna P. Palliative biliary stents for obstructing pancreatic
carcinoma. Cochrane Database Syst Rev 2006;(2):CD004200. Review.
Moss AC, Morris E, Leyden J, et al. Malignant distal biliary obstruction: a systematic
review and meta-analysis of endoscopic and surgical bypass results. Cancer Treat
Rev 2007;33:213–221.
Petrov MS, van Santvoort HC, Besselink MG, et al. Early endoscopic retrograde
cholangiopancreatography versus conservative management in acute biliary
pancreatitis without cholangitis: a metaanalysis of randomized trials. Ann Surg
2008;247:250–257.
Puli SR, Reddy JB, Bechtold ML, et al. EUS-guided celiac plexus neurolysis for pain due
to chronic pancreatitis or pancreatic cancer pain: a metaanalysis and systematic
review. Dig Dis Sci 2009;54:2330–2337.
van der Gaag NA, Rauws EA, van Eijck CH, et al. Preoperative biliary drainage for
cancer of the head of the pancreas. N Engl J Med 2010;362:129–137.
van Santvoort HC, Besselink MG, Bakker OJ, et al; Dutch Pancreatitis Study Group.
A step-up approach or open necrosectomy for necrotizing pancreatitis. N Engl J
Med 2010;362(16):1491–1502.
Varadarajulu S, Christein JD, Tamhane A, et al. Prospective randomized trial compar-
ing EUS and EGD for transmural drainage of pancreatic pseudocysts. Gastrointest
Endosc 2008;68:1102–1111.
Wang Q, Gurusamy KS, Lin H, et al. Preoperative biliary drainage for obstructive jaun-
dice. Cochrane Database Syst Rev 2008:CD005444.
Yim HB, Jacobson BC, Saltzman JR, et al. Clinical outcome of the use of enteral stents
for palliation of patients with malignant upper GI obstruction. Gastrointest
Endosc 2001;53:329.
002086391.INDD 408 7/14/2014 8:29:38 PM

33 HPB Trauma
Chad G. Ball
Injuries to the liver, extrahepatic biliary tree, and pancreas are often deadly
and always challenging. This area is also commonly referred to as the “sur-
gical soul.” They will engage all of your senses, test your skills, and demand
great teamwork from you and your colleagues.
HEPATIC INJURIES
Although textbooks are often filled with a complex hierarchy of operative
interventions and maneuvers for treating hepatic trauma, the vast majority
of liver injuries are treated nonoperatively. This process involves diagnosis
with cross-sectional imaging (CT), serial clinical observation, and labora-
tory (hemoglobin, white blood cell count [WBC], and liver function tests/
enzymes) monitoring. This algorithm allows the clinician to successfully
treat and predict both the initial injury and potential complications such as
delayed/ongoing hemorrhage and interval formation of a biloma. Clearly,
any patient who presents with hypotension and/or peritonitis (i.e., concur-
rent injuries) requires emergent operative therapy.
Management
The dominant challenge with hepatic trauma is management of the hemo-
dynamically unstable patient with ongoing hemorrhage from a high-grade
liver injury. These patients often present in physiologic extremis and therefore
require damage control resuscitation techniques. Early recognition of their crit-
ical condition, as well as immediate hemorrhage control, is essential to survival.
Patients with major injury caused by blunt trauma or right upper
quadrant penetrating mechanisms must undergo an immediate F.A.S.T.
(i.e., Focused Assessment with Sonography for Trauma) examination in the
trauma bay to confirm the presence of large-volume intraperitoneal fluid.
This exam is repeatable and can be used to reevaluate patients in urban cen-
ters who present immediately following their injuries. Massive transfusion
protocols as part of a damage control resuscitation must be initiated early
during the patient assessment process. If the patients rapidly stabilize their
hemodynamics, they should undergo an emergency CT scan of the torso
(blunt and gunshot). If they remain clinically unstable, however, patients
must be transferred to the operating theater without delay. Hemorrhage con-
trol is the dominant driver in these patients. Collateral issues such as optimal
intravenous access, imaging of other areas (brain, spine, bones), and fracture
fixation are important but, nevertheless, secondary priorities. In summary:
1. In hemodynamically stable patients without CT evidence of a hepatic
arterial blush or other reasons to proceed to the operating theater,
admission and close observation are warranted.
409
0002086392.INDD 409 7/14/2014 8:35:25 PM

2. In hemodynamically stable patients with a hepatic arterial extravasa-

tion/blush, immediate transfer to the interventional angiography suite
(or hybrid O.R.) is mandated for hepatic angiography and/or portography
with selective embolization. Autologous clot or absorbable embolization
medium is optimal.
3. In patients with persistent hemodynamic instability, immediate transfer
to the operating theater for laparotomy is essential. Delays will lead to
loss of life.
Operative Care and Technical Tips

The patient should be rapidly prepared and draped with available access
from the neck to the knees. Vascular instruments and balloons must be
open and ready within the theater. A midline laparotomy from xiphoid
process to pubic bone should be performed with three passes of a sharp
scalpel. The peritoneal cavity should be packed in its entirety with laparot-
omy sponges for patients with blunt liver injuries. The falciform ligament
should be left intact to provide a medial wall against which to improve
medial packing pressure. The right upper quadrant (and midline vascular
structures) should be evaluated prior to any intraperitoneal packing for
penetrating injuries. If hemorrhage continues, an early Pringle maneuver
(clamping of the porta hepatis with a vascular clamp) is mandated. This is
both diagnostic and potentially therapeutic. If bleeding continues despite
application of a Pringle clamp, a retrohepatic inferior vena cava (IVC) or
hepatic venous injury is likely. Critically injured patients in physiologic
extremis do not tolerate extended Pringle maneuvers to the same extent
as patients with hepatic tumors undergoing elective hepatic resection.
Forty minutes represents the upper limit of viability. If the liver hemor-
rhage responds to packing (which describes the vast majority of cases
85% to 99% depending on the series) but continues to hemorrhage when
unpacking is completed, the patient should be repacked and transferred
to the ICU with an open abdomen once damage control of concurrent
injuries is complete. Cover the liver with a plastic layer of sterile x-ray cas-
sette material to avoid capsular trauma upon eventual unpacking. Topical
hemostatic agents are also helpful. If control of the liver hemorrhage is
dependent on maintenance of a Pringle maneuver despite packing (i.e.,
the retrohepatic IVC hematoma has ruptured), call for senior assistance,
mobilize the right lobe, and suture the IVC or hepatic veins with 4-0
Prolene on SH needles or 5-0 Prolene on RB1 needles (assuming that a
replaced left hepatic artery is not the source of inflow occlusion failure).
These patients may also require total vascular exclusion (TVE) of the liver
(complete occlusion of the infrahepatic IVC, suprahepatic IVC, porta
hepatis (Pringle maneuver), as well as an aortic cross-clamp within the
abdomen). If TVE is pursued without concurrent clamping of the aorta,
the patient will arrest due to a lack of coronary perfusion. I prefer to obtain
suprahepatic IVC control within the abdomen in patients with a normal
length of IVC inferior to the diaphragm. An alternate option includes
access of the IVC within the pericardium itself (i.e., as it enters the heart).
This 2-cm length of IVC is easily accessible by opening the pericardial sac
after dividing the diaphragm. Alternatively, the suprahepatic IVC can also
be accessed from the thorax if a thoracotomy has already been performed.
Veno–veno bypass is also a theoretical option but rarely used due to a lack
of transplant training for most trauma/general surgeons.
In the case of central hepatic gunshot wounds or deep central lac-
erations where access and exposure are difficult, ongoing hemorrhage
should be stopped with balloon occlusion. Either a Blakemore esophageal
balloon or variant (red rubber catheter with overlying Penrose drain and
002086392.INDD 410 7/14/2014 8:35:25 PM

Chapter 33 / HPB Trauma 411
two silk occlusion ties) is exceptional at stopping ongoing bleeding at the

bottom of deep central hepatic injury tracts (including retrohepatic IVC
injuries). These should be deflated approximately 72 hours after the ini-
tial placement. If hemorrhage continues, they should be left in situ for 2
to 3 additional days. Although unusual, penetrating injuries to the hepatic
artery may require ligation (assuming the portal vein is intact). Portal vein
injuries should ideally be repaired with 5-0 or 6-0 Prolene once control is
obtained. Clamps above and below the injury are essential for visualiza-
tion. Alternate damage control options for portal vein injury include TIVS
(temporary intravascular shunt) with a small chest tube conduit and liga-
tion (assuming the hepatic artery is intact). Return to the operating suite
in patients with packed abdomens should occur in 48 to 72 hours (assum-
ing hypothermia, coagulopathy, and acidosis are corrected). If an atrial–
caval shunt is contemplated, two experienced surgical teams (one for the
chest and one for the abdomen) are essential. The decision to pursue this
shunt must be made early in the exploration process as they rarely result in
patient salvage.

Complications following hepatic trauma include both early (bleeding and
abdominal compartment syndrome) and late (infected hematoma, bile
leak, liver failure, and associated ileus) diagnoses. Delayed laparoscopic
washout with copious irrigation and insertion of a closed suction drain-
age catheter is particularly helpful in patients who have large collections
of infected hematoma and/or bilomas. This is decidedly more common
in high-grade hepatic injuries (8% to 10% risk of biloma in grade III to V
injuries). If a bile leak persists, then placement of an intrabiliary stent via
endoscopic retrograde cholangiopancreatography (ERCP) is helpful in
lowering biliary pressure and encouraging closure of the leak. As a result,
successful “nonoperative” therapy requires access to percutaneous drain-
age, laparoscopy, ERCP, and/or angiography in selected cases. It should
also be noted that gunshot wounds with right upper quadrant trajectories
(i.e., thoracoabdominal) can be treated nonoperatively in the context of
normal hemodynamics and the absence of peritonitis. Concurrent injuries
to the duodenum and transverse colon must be ruled out however (i.e., triple-
contrast CT). Furthermore, one must also monitor for the development of
a bronchobiliary fistula. Most bronchobiliary fistula can be successfully
treated with a biliary stent and appropriate drainage. Hepatic failure typi-
cally occurs in response to shock liver and/or inflow occlusion (Pringle)
required at the time of the operative procedure. Standard supportive care
for hepatic failure is indicated.
Outcomes and Follow-up

Planned surveillance cross-sectional imaging (e.g., CT) is not required for
major liver injuries (as opposed to splenic injuries). Repeat imaging should
be based on any deterioration in laboratory tests or patient symptoms. The
appropriate delay in time to return to physical/combat sports is debatable
following major hepatic injury. Despite a known 4- to 8-week hypertrophy
response following elective hepatic resection, the time to regeneration and
complete organ healing is unclear in the context of hepatic injuries.
Conclusion
If diagnosis and therapy are rapid, patients who present in physiologic
extremis as a result of major hepatic hemorrhage have a good chance of
survival in the context of a prolonged hospital stay. Complications must be
managed appropriately and without delay.
0002086392.INDD 411 7/14/2014 8:35:25 PM

EXTRAHEPATIC BILIARY TRACT INJURIES

Extrahepatic biliary tree (common bile duct and gallbladder) injuries are
incredibly uncommon. They are also relatively simple to treat relative to
the context of nonelective (i.e., laparoscopic cholecystectomy) injuries.
Diagnosis typically occurs within the operating theater at the time of
urgent exploration for concurrent injuries. Trauma to the biliary tree is
suspected upon identification of bile within the peritoneal cavity. If the
diagnosis is delayed, a biloma will form in the setting of a typically sterile
field. As a result, patient symptoms will consist of nausea, mild right upper
quadrant discomfort, and often an ileus. The WBC and bilirubin levels may
also be elevated. These patients require identification of the biloma with
either ultrasound or CT, in addition to subsequent percutaneous drainage
and cholangiography. In scenarios of very small partial-thickness common
bile duct injuries, placement of an intrabiliary stent via ERCP may be suf-
ficient. With any significant injury, however, immediate control of sepsis
and subsequent, appropriately timed exploration by an HPB surgeon is
warranted.
Management
In the setting of a hemodynamically stable patient with a delayed diagnosis
of an extrahepatic biliary tract injury, complete cholangiography is essen-
tial prior to exploration. An experienced colleague and/or team approach
is also crucial.

Any injury to the gallbladder is an absolute indication for cholecystec-
tomy. Although primary repair is occasionally described in very large
series, it is almost never indicated. A full-thickness common bile duct
injury is generally an indication for a standard Roux-en-Y hepaticojeju-
nostomy to minimize the long-term risk of anastomotic strictures. In the
setting of small, partial-thickness sharp (i.e., noncautery and nongun-
shot) injuries, primary repair with T-tube drainage/control may also be
a viable option.
Postoperative Management, Complications, and Follow-up

Although most high-volume hepatobiliary surgeons do not utilize closed
suction drainage for their hepaticojejunostomies, injury in the context of
patients with additional trauma and physiologic stressors may provide an
indication for drainage in some scenarios. The dominant long-term poten-
tial complication remains stricture of the biliary anastomosis. Stricture
is particularly plausible in the context of hepaticojejunostomies required
for very youthful patients with a long life expectancy. It also mandates a
detailed discussion with the patient prior to discharge (i.e., risks and symp-
toms associated with stricture—cholangitis, jaundice). These patients are
often amenable to dilations of their anastomoses with either an endoscopic
or a percutaneous approach.
Conclusion
Injury to the extrahepatic biliary tree is unusual. Cholecystectomy is indi-
cated for any trauma to the gallbladder. Full-thickness common bile duct
injuries require a Roux-en-Y hepaticojejunostomy, whereas minor partial-
thickness injuries can occasionally be treated with primary repair and
decompression.
002086392.INDD 412 7/14/2014 8:35:25 PM

Chapter 33 / HPB Trauma 413
PANCREAS INJURIES
As with hepatic trauma, injuries to the pancreas can be challenging to
even the most experienced trauma/general surgeon. Pancreatic injuries are
divided into injuries of the left (body/tail) and right (head) organ. Clearly,
injuries to the right pancreas place adjunctive structures (i.e., portal vein,
IVC, aorta, duodenum, bile duct) at risk as well.
Management
Absolute indications for operative exploration of a suspected injury to the
pancreas include (1) hypotension, (2) diffuse peritonitis, and (3) obvious
full-thickness organ disruption in the pancreatic neck/body/tail on cross-
sectional imaging.

All pancreatic hematomas should be opened and explored to define the
integrity of the pancreatic capsule. All segments and surfaces of the pan-
creas can be directly inspected by utilizing standard exposure maneuvers
(Kocher maneuver, division of ligament of Treitz, entry into the lesser sac,
medial mobilization of the tail of pancreas and spleen). If a disruption in
the pancreatic capsule is noted (grade I or II), a closed suction drain must
be placed adjacent to the injury. If a pancreatic laceration/hole is noted
in the setting of an intact main pancreatic duct, a viable omental plug or
patch should be combined with closed suction drainage. In general, inju-
ries with disruption of the main pancreatic duct at any location to the left
of the superior mesenteric vein/portal vein (SMV/PV) should undergo a
distal pancreatectomy (grade III). Endoscopic (ERCP)-based pancreatic
duct stenting across a truly disrupted main pancreatic duct fails almost
uniformly. Similarly, a Roux-en-Y pancreaticojejunostomy to the pancreatic
segment distal to the transaction is typically fraught with complications
and fistula due to the soft texture of an otherwise normal gland.
Defining the integrity of the main pancreatic duct can be challeng-
ing. Ultrasonography has become incredibly valuable for surgeons who are
trained to evaluate pancreatic gland anatomy. In the absence of this skill,
however, intraoperative cholecystocholangiography remains the best alter-
native option. Utilize a combination cocktail of both methylene blue and
radioopaque contrast material (± fentanyl) when obtaining images.
Injuries to the right pancreas (head) with main duct disruption
typically require a pancreatoduodenectomy. In the context of damage
control resuscitation due to massive hemorrhage, the initial surgery does
not typically allow for either a completed resection or reconstruction. As
a result, the “trauma Whipple” is usually performed in two stages. The
second procedure (completion of resection and reconstruction) should
be performed within 24 hours. More specifically, if reexploration is
delayed beyond 24 hours, the viscera become so edematous that recon-
struction is extremely challenging due to limited mobility of the pan-
creatobiliary limb and difficult, sloppy anastomoses. It should also be
noted that most combined pancreaticoduodenal injuries do not require a
Whipple procedure. In these patients (i.e., with ampullary, common bile,
and pancreatic duct integrity), procedures such as pyloric exclusion with
gastrojejunostomy or the Cali triple tube decompression are superior
options. It must be reemphasized that the “bailout” maneuver for pan-
creatic head trauma remains drainage (closed suction or negative suc-
tion dressing) with early reexploration in conjunction with experienced
pancreatic surgeons.
0002086392.INDD 413 7/14/2014 8:35:25 PM


Regardless of the technique, up to 25% of distal pancreatectomies will
generate a pancreatic leak and subsequent fistula. This fistula rate may be
higher in the setting of trauma due to the impact of concurrent injuries
and physiologic stress/exhaustion. Soft texture, contaminated fields, and
bruised pancreas glands make surgery of the pancreas in the injured popu-
lation challenging and fraught with complications. Wrapping the pancreas
stump with viable omentum may be helpful. Octreotide has not been reli-
ably shown to reduce the time to closure of pancreatic fistulas in injured
patients. In addition to these early consequences, delayed endocrine insuf-
ficiency (i.e., diabetes) is also a concern. It is clear that upon removal of
60+% of pancreatic tissue in young trauma patients, approximately half will
go on to achieve glucose intolerance, and up to 25% will develop diabetes.
Conclusion
Injuries to the pancreatic head typically involve catastrophic concurrent
injuries to the surrounding organs and vasculature. Damage control scenar-
ios demand drainage once ongoing hemorrhage is stopped. Injury to the left
pancreas most commonly requires a distal pancreatectomy. Maintaining
control of a postoperative pancreatic fistula is essential at all times.
Suggested Readings
Ball CG, Dixon E, Kirkpatrick AW, et al. A decade of experience with injuries to the
gallbladder. J Trauma Manag Outcomes 2010;4:3.
Ball CG, Wyrzykowski AD, Nicholas JM, et al. A decade’s experience with balloon tam-
ponade for the emergency control of hemorrhage. J Trauma 2011;70:330–333.
Chinnery GE, Krige JE, Kotze UK, et al. Surgical management and outcome of civilian
gunshot injuries to the pancreas. Br J Surg 2012;99.
Feliciano DV. Biliary injuries as a result of blunt and penetrating trauma. Surg Clin
North Am 1994;74:897–907.
Kozar RA, Feliciano DV, Moore EE, et al. Western Trauma Association/critical deci-
sions in trauma: operative management of adult blunt hepatic trauma. J Trauma
2011;71:1–5.
Pachter HL. Prometheus bound: evolution in the management of hepatic trauma—
from myth to reality. J Trauma 2012;72:321–329.
Pachter HL, Feliciano DV. Complex hepatic injuries. Surg Clin North Am 1996;76:
763–782.
Subramanian A, Dente CJ, Feliciano DV. Management of pancreatic trauma in the
modern era. Surg Clin North Am 2007;87:1515–1532.
002086392.INDD 414 7/14/2014 8:35:25 PM

34 Interventional Radiology
Support of HPB Surgery
Robert L. King and Matthew S. Johnson
Tremendous advances in interventional radiology (IR) technique have

benefitted the interventional radiologist and surgeon alike. A close relation-
ship with IR physicians is crucially important to ideal HPB surgery practice
and management of complex HPB patients. Multiple IR techniques play
important roles in preoperative diagnosis, patient optimization (i.e., biliary
drainage, portal vein embolization), definitive treatment (e.g., percutane-
ous ablation, transarterial treatment of liver tumors), and rescue of HPB
patients from postoperative complications (e.g., treatment of pseudoaneu-
rysms, drainage of intra-abdominal abscess).
The goal of this chapter is to familiarize HPB practitioners with spe-
cific IR techniques: percutaneous biliary stenting, transjugular intrahepatic
portosystemic shunting (TIPS), portal vein embolization (PVE), percutane-
ous radiofrequency ablation (RFA), transarterial treatment of liver tumors,
percutaneous drainage of intra-abdominal abscess, and transarterial treat-
ment of pseudoaneurysms (PSAs).
PERCUTANEOUS TRANSHEPATIC CHOLANGIOGRAPHY AND BILIARY

INTERVENTION IN THE MANAGEMENT OF BILIARY OBSTRUCTION
As endoscopic techniques have become more common in cases of bili-
ary obstruction, percutaneous biliary interventions have decreased in
frequency. Percutaneous biliary intervention continues to be useful when
endoscopic techniques have failed or in instances where they are unlikely to
succeed. Given the complexity and risks associated with the procedure, the
decision to move forward with the percutaneous procedure is usually made
with multidisciplinary expertise, including HPB surgeons, interventional
radiologists, and endoscopists.
Biliary obstruction may be caused by benign or malignant pathol-
ogy, and obstructive jaundice can be confirmed with ultrasound, com
puted tomography (CT), or magnetic resonance cholangiopancreatography
(MRCP). Ultrasound may demonstrate the presence of intrahepatic or
extrahepatic biliary ductal dilation and may demonstrate the level of
obstruction. Contrast-enhanced CT of the abdomen is more likely to
identify the level and nature of obstruction. While MRCP provides excel-
lent three-dimensional imaging, this study should be used as an adjunct
to contrast-enhanced CT given its limited ability to demonstrate surgical
clips, hardware, and other anatomic landmarks. In cases where bile leak
is suspected, radionuclide hepatobiliary scintigraphy may be performed.
Indications and contraindications are shown in Table 34.1.
Biliary intervention can be performed with conscious sedation, deep
sedation, or general anesthesia. Patients should be made NPO with clear
liquids only 6 hours prior to the procedure. Patients should have intrave-
nous access and be hydrated, as manipulation of the biliary tree can cause
a patient to become septic and hypotensive, requiring rapid resuscitation.
415
0002100651.INDD 415 7/14/2014 8:37:38 PM

TABLE
Indications and Contraindications to Percutaneous Biliary
34.1
Indications
Intervention
1. Precisely define biliary anatomy

2. Drain obstructed bile ducts
3. Obtain tissue or bile for diagnosis
4. Manage benign strictures
5. Manage intrahepatic or extrahepatic stone disease
6. Manage biliary malignancies
Contraindications
Absolute
1. Uncorrectable coagulopathy
2. No available percutaneous access to the biliary system (e.g., colonic inter-
position or very high left liver remnant)
Relative
1. Ascites (consider paracentesis before or during procedure)
2. Intrahepatic processes that cause multilevel obstruction not amenable to
percutaneous treatment (e.g., polycystic liver disease or widespread intra-
hepatic metastases)
Per Society of Interventional Radiology (SIR) guidelines, prophylactic anti-

biotics should be administered immediately prior to instrumenting the
obstructed biliary tree. If culture data are available, antibiotic choice should
be tailored to specific organisms. More recently, and in patients with previ-
ous biliary instrumentation, the incidence of resistant organism coloniza-
tion in the bile is very high.
For a left biliary duct puncture, a subxiphoid approach is used. For
a right puncture, a lateral approach is used with the cranial/caudal posi-
tion as low as possible. Ideally, this will be below the 10th rib to avoid a
transpleural puncture. The access needle is then passed through the
hepatic parenchyma under fluoroscopic guidance. Contrast is injected
through the needle as the needle is slowly retracted until a bile duct is
opacified. Once a bile duct is identified, a cholangiogram is then performed
using gentle short injections of contrast until adequate filling of the ducts
is obtained. (It is not always possible to opacify central ducts from the ini-
tial access. Instead, a “regional” cholangiogram is performed to assess the
local ducts and the access site.) If the initial access is acceptable—that is,
the duct is entered as peripherally as possible and the central angles are
favorable for placement of a catheter—it may be used for placement of a
6- to 8-Fr sheath. Alternatively, a second access may be obtained by choos-
ing an acceptable peripheral duct, which has been opacified from the origi-
nal injection. It may be necessary to intermittently inject contrast though
the initial access to keep the ducts adequately opacified. A flexible guide-
wire is then advanced through the access needle into a central duct. The
needle is removed and replaced with a 6- to 8-Fr sheath over the guidewire.
A complete cholangiogram can then be performed through this sheath or
through a 4- to 5-Fr catheter, if the sheath tip is not central enough to obtain
adequate visualization of the central ducts. The obstruction is then crossed
with a combination of a hydrophilic wire and 4- to 5-Fr catheter. The same
wire/catheter combination is then advanced into the proximal jejunum.
The hydrophilic wire is then exchanged for a nonhydrophilic stiff wire, the
tract is then dilated to the appropriate size, and an internal/external biliary
stent is then placed with the tip positioned as distal in the bowel as possible
002100651.INDD 416 7/14/2014 8:37:38 PM

Chapter 34 / Interventional Radiology Support of HPB Surgery 417
while maintaining the peripheral catheter side hole at the ductal site entry
point. This position ensures that the ducts peripheral to the puncture site
are not excluded from drainage. If the obstruction cannot be crossed, an
external biliary drain is placed with its tip as central as possible. The cath-
eter is then sutured to the skin and placed to an external gravity drainage
bag. The internal/external stent can usually be capped the following day
and may remain capped as long the patient is afebrile and without other
signs of cholangitis.
Malignant strictures can be treated as above, or an indwelling stent
may be placed in cases of palliation. The lesion is crossed using conventional
techniques. Occasionally, cholangioplasty may be performed to facilitate
crossing of the lesion with the stent. Self-expanding bare metal or covered
stents may be utilized. Once the stent is in place, a repeat cholangiogram
is performed to ensure free flow of contrast across the stent. When this has
been confirmed, the parenchymal access tract is coil embolized.
Technical Tips
1. Puncture the biliary system as peripheral as possible to avoid ductal
exclusion and injury of larger central vessels. (Fig. 34.1 shows large intra-
hepatic hematoma from laceration of hepatic artery.)
2. During pullback injection, small “puff ” of contrast should be introduced.
Contrast in the portal vein and hepatic artery will head to the periphery
FIGURE 34.1 Coronal CT showing large intrahepatic hematoma (long arrow)

from inadvertent hepatic artery laceration during percutaneous transhepatic
cholangiography (PTC) (short arrows) placement.
0002100651.INDD 417 7/14/2014 8:37:39 PM

of the liver and clear completely. The hepatic vein will clear centrally
toward the heart.
3. Overinjecting a dilated infected biliary system can lead to sepsis, shock,
and death.
4. If the wire does not initially advance smoothly, the entry angle of the duct
may be too acute or the needle may be just outside the duct. Pulling back
and repeating a more peripheral puncture will likely be helpful.
5. A short nitinol wire, increased fluoro magnification, and twirling move-
ments of the wire may be beneficial.
6. Advancing a 6-Fr 25-cm-long sheath immediately proximal to the
obstruction will allow more forward force to be placed on the coaxial
catheters and guidewires.
7. The duct distal to the obstruction can be punctured, and a snare can be
placed allowing through and through access.
Postprocedural Management
The patient should have his or her vital signs checked frequently immedi-
ately after the procedure. While procedure-related complications follow-
ing biliary catheter placement are infrequent, bleeding from injury to the
hepatic artery or portal vein is possible. This usually manifests as bleeding
through or around the catheter, but would also be suggested by tachycardia
with or without hypotension. Cholangiography might demonstrate com-
munication between the catheter tract and blood vessel. Portal bleeding
is usually easily treated by replacement of the indwelling catheter with a
bigger catheter and/or one advanced more centrally such that its side holes
no longer communicate with the portal vein. Hepatic arterial bleeding is
usually treated with embolization of the injured artery. Pneumothorax and
the equally uncommon bilious pleural effusion are rare complications; both
are treated by catheter drainage of the pleural space.
The external or internal/external biliary catheter should undergo rou-
tine catheter care including daily site cleaning and dressing changes. Biliary
catheters, whether external drains or internal/external catheters, should be
flushed: Our standard procedure is to flush catheters with 10 mL of sterile
saline twice a day. Bile leak around the catheter, persistent pericatheter or
right upper quadrant pain, and/or difficulty flushing the catheter are all
indications for cholangiography and possible catheter replacement.
Patients are observed overnight for symptomatic management and
to allow for early detection of delayed procedural complications. At our
institution, patients with benign bile duct strictures are normally man-
aged with 12 months of stenting. Once the initial stent has been placed
(usually 10 to 12 Fr), it is serially upsized every few months to 20 Fr as tol-
erated by the patient. The 12-month timeframe starts once the 20-Fr stent
has been placed. The stent is exchanged every 3 months during this time. At
12 months, serum liver chemistry tests and cholangiogram are performed.
If contrast flows freely through the treated stricture, a “clinical trial” is initi-
ated. This involves placing a new catheter such that the tip is proximal to the
previous stricture. The catheter is then capped for 2 weeks, at which time
the patient returns for a repeat cholangiogram and serum liver chemistry
evaluation. If the liver chemistry tests are stable, the cholangiogram shows
resolved stricture, and the patient is doing well clinically (e.g., no fevers or
signs of cholangitis), the catheter is removed. An alternative to the clinical
trial to test the integrity of a treated bile duct stricture is the Whittaker test,
which involves manometry measurement of bile duct pressure proximal
and distal to the stricture. Patients with “definitively” treated benign bile
duct strictures must have long-term annual follow-up of liver chemistry
tests with their primary physician, gastroenterologist, or HPB surgeon.
002100651.INDD 418 7/14/2014 8:37:39 PM

Immediate Complications
1. Bleeding (subcapsular/intraparenchymal/peritoneal/pleural/
biliary—hemobilia)
2. Infection (cholangitis/sepsis)
3. Bile duct perforation
4. Stent migration/malposition
Delayed Complications
1. Infection (pancreatitis/cholecystitis/cholangitis)
2. Pancreatitis
3. Bile leak in the peritoneum or pleural space
4. Stent occlusion, migration, kinking, or dislodgement
5. Biloma
The average rate of major complications is about 2%, although complica-
tion rates can vary from 4% to 25% depending on the institution and opera-
tor experience. The majority of these complications can be treated by the
interventionalist. Pleural transgression is usually treated with a chest tube.
A patient who is bleeding should be treated with standard resuscitative
measures. In addition, coagulopathy secondary to jaundice must be cor-
rected (administration of vitamin K and replacing plasma clotting factors).
Bleeding in and around a catheter is usually due to a side hole within a
blood vessel that was crossed. Advancing the catheter will stop the major-
ity of the bleeding that is venous. If the patient continues to bleed, an arte-
riogram should be performed with the biliary catheter removed over a wire,
maintaining access. Coil embolization distal and proximal to the bleeding
arterial site is therapeutic. Infectious cholangitis should be treated with
7 to 10 days of antibiotic coverage for biliary/gastrointestinal (GI) bacte-
ria. Additionally, a cholangiogram should be performed to ensure the cath-
eter is functioning properly. Sepsis or septic shock requires ICU admission.
Insertion-site cellulitis may suggest misplaced side holes and should be
managed with cholangiography with catheter replacement, if indicated. If
cellulitis is present with a well-positioned patent catheter, then antibiot-
ics should be administered. Biliary catheter obstruction usually requires
reevaluation with repeat cholangiogram and replacement. Biliary catheter
dislodgement should prompt urgent evaluation as newly created tracts may
close within hours. Stent occlusion can be treated with recanalization and
cholangioplasty, or new stent placement.
Outcomes/Follow-up
Effective biliary drainage is established in close to 100% of bile ducts
accessed in appropriately selected patients. Percutaneous drainage can
successfully treat between 70% and 100% of iatrogenic bile duct injuries.
Reported technical success rate for nontransplant bile duct strictures
approaches 100% with transplant stenosis having a more modest success
rate of 45% to 80%. The covered stent patency for biliary obstruction is 90%
to 95%, 76% to 92%, and 76% to 85% at 3, 6, and 12 months, respectively.
TRANSJUGULAR INTRAHEPATIC PORTOSYSTEMIC SHUNT

A TIPS is a percutaneous technique used to treat portal hypertension and
its complications. A TIPS is placed to facilitate reduction of the portal
venous pressure by placing a decompressive stent between a hepatic vein
and an intrahepatic portal vein (Fig. 34.2 shows TIPS). Since its inception
over 20 years ago, more than 1,000 patients have been enrolled in multiple
controlled trials studying efficacy and safety of the procedure. Furthermore,
0002100651.INDD 419 7/14/2014 8:37:39 PM

FIGURE 34.2 (A-C). Schematic of TIPS. (From Mulholland MW, Lillemoe
KD, Doherty GM, et al. Greenfield’s surgery: scientific principles & practice,
002100651.INDD 420 7/14/2014 8:37:41 PM

more than 1,000 papers have been published allowing for the development
of guidelines to define the role of TIPS in the management of portal
hypertension.
Cirrhosis is common throughout the world and may lead to ascites
formation, variceal bleeding, and hepatic encephalopathy. Of these compli-
cations, variceal bleeding is the most common indication for TIPS. In fact,
the mortality rate within 6 weeks of bleeding is 30%, which can most com-
monly be attributed to uncontrolled or early recurrent bleeding.
The model for end-stage liver disease (MELD) score was developed
specifically to predict post-TIPS outcomes. This formula, which takes into
account the creatinine, bilirubin, and international normalized ratio (INR),
is an accurate predictor of 30-day mortality. If the MELD score is 1 to 10, the
30-day mortality is 3.7%. If the MELD is greater than 24, the 30-day mortality
rises to 60%.
As the name indicates, TIPS is placed via a right internal jugular vein
approach. It may be performed with moderate sedation or general anesthe-
sia. The type of stent that is placed is operator dependent. Current evidence
suggests that a PTFE-covered stent (Viatorr) may have the best long-term
outcomes.
Technical success rates are greater than 95%. Hemodynamic success
rates are greater than 95%. Clinical success rates are measured in terms of
survival, control of bleeding, or ascites. Control of bleeding is achieved in
90%, and improved ascites control is seen in 60% to 85%.
Complications include stent malposition, hemobilia, radiation skin
burn, hepatic infarction, acute TIPS thrombosis, stent stenosis (usually
due to neointimal hyperplasia), liver failure, encephalopathy, heart failure,
recurrent portal hypertension, and recurrent bleeding. Indications and
contraindications to TIPS are shown in Table 34.2.
TABLE
34.2
Indications
Indications and Contraindications to TIPS
1. Secondary prevention of variceal bleeding

2. Refractory acutely bleeding varices
3. Refractory cirrhotic ascites
4. Portal hypertensive gastropathy
5. Gastric antral vascular ectasia
6. Refractory hepatic pleural effusion
7. Hepatorenal syndrome (type 1 or 2)
8. Budd-Chiari syndrome
9. Venoocclusive disease
10. Hepatopulmonary syndrome
Contraindications
Absolute
1. Severe or rapidly progressive liver failure
2. Severe or uncontrolled encephalopathy
3. Heart failure
Relative
1. Severe uncorrectable coagulopathy
2. Uncontrolled sepsis
3. Unrelieved biliary obstruction
4. Extensive primary or metastatic hepatic malignancy
5. Polycystic liver disease
0002100651.INDD 421 7/14/2014 8:37:41 PM

Preoperative Evaluation
Contrast-enhanced CT is necessary to evaluate for underlying malignancy,
anatomy, and patency of the hepatic and portal veins. Alternatively, a
Doppler ultrasound of the liver may be performed to evaluate vessel patency.
Appropriate lab work includes liver function tests and alpha-fetoprotein.
Postoperative Care
Admit for 24-hour observation, if the procedure is done electively. When
TIPS is placed emergently, the patient is normally admitted to the ICU. Post-
TIPS ultrasound should be ordered the following day if a bare metal stent
is placed and in 1 to 2 weeks if a Viatorr (PTFE-covered) stent is used. This
test is used as a baseline for future follow-up of the shunt. Changes in blood
flow velocity through the stent predict potential shunt issues.
PREOPERATIVE PORTAL VEIN EMBOLIZATION

Liver resection often times is the only option that may confer long-term sur-
vival in patients with primary and secondary liver malignancies. However,
patients with limited liver remnant volumes are at increased risk of post-
operative liver failure and death. Preoperative PVE has emerged as a use-
ful preoperative technique to stimulate hypertrophy of the future remnant
liver (FRL) to facilitate the possibility of a major hepatectomy.
Rous and Larimore were the first to demonstrate the regenerative
capacity of the liver following portal vein occlusion in the 1920s. In 1961, the
first human portal vein ligation was performed during a two-stage hepatec-
tomy. The first preoperative PVE in humans was performed in 1986. Since
then, PVE has been shown to be safe and effective in inducing FRL hyper-
trophy in preparation for surgical resection.
Today, percutaneous transhepatic PVE is the technique of choice. PVE
can be performed from an ipsilateral or contralateral approach. The choice
of approach is usually based on the tumor burden within the liver, the need
for segment 4 embolization, and operator preference. Many options for
embolics exist, including Gelfoam, coils, Amplatzer plugs, polyvinyl alcohol
particles, n-butyl cyanoacrylate and Lipiodol, or fibrin glue.
The technical success rate of PVE is 99%, and the clinical success
rate is 96%. Few patients are unsuitable for hepatic resection because
of inadequate hypertrophy. Regardless of approach, complications rates
range from 9% to 15%. Complications include subcapsular hematoma,
hemobilia, AV fistula, pseudoaneurysm, arterioportal shunt, nontarget
embolization, portal vein thrombosis (of the FRL), portal hypertension,
and sepsis.
Patient Selection
1. Patients with primary or metastatic liver disease, who are otherwise
hepatic resection candidates, except for the following:
a. Cirrhosis and/or advanced fibrosis and an FLR/total liver volume
(TLV) < 40%
b. Extensive chemotherapy and an FLR/TLV < 30%
c. Normal underlying liver and an FLR/TLV < 20%
2. Patients with diabetes mellitus but without underlying liver disease may
benefit from PVE, as the magnitude of postresection liver hypertrophy is
usually less in these patients.
3. Patients undergoing complex hepatectomy with concomitant extrahe-
patic surgery, particularly pancreatectomy. In this group, studies have
shown that hepatic regeneration is inversely proportional to the extent
of pancreatectomy.
002100651.INDD 422 7/14/2014 8:37:41 PM

FIGURE 34.3 Preoperative volumetric analysis of a patient with right intrahepatic cholan-
giocarcinoma (arrow). Planned operation is right hepatectomy; note FLR outlined in blue.
Preoperative Workup
Volumetric 3D contrast-enhanced CT is essential for planning hepatic
resection and calculating TLV and FLR. These measurements are easily cal-
culated by dedicated abdominal radiologist (Fig. 34.3).
Postoperative Care
Admit for 24-hour observation. Patients generally do not develop postem-
bolization syndrome as PVE leads to apoptosis as opposed to necrosis.
Hydrate vigorously until the patient tolerates oral hydration. Symptomatic
treatment of pain, fever, and nausea. Repeat CT is typically performed in
3 to 4 weeks to evaluate FLR hypertrophy and recalculate volumetrics.
RADIOFREQUENCY ABLATION
Image-guided percutaneous techniques for local tumor ablation offer
effective treatment of select hepatic malignancies including hepatocellu-
lar carcinoma (HCC) and colorectal metastases. Although many ablation
techniques exist, RFA is the primary ablative technique used currently.
Traditionally, ultrasound has been the imaging modality of choice for guid-
ance. However, hepatic lesions are not always visible by this technique.
In cases where lesions are not visible on gray-scale ultrasound, contrast-
enhanced ultrasound, CT, and magnetic resonance can be utilized.
The goal of RFA is to induce thermal injury to the tumor through elec-
tromagnetic energy deposition. In fact, the patient is part of a closed loop
circuit that includes the RF generator, the electrode needle, and a large dis-
persive electrode. An alternating electric field is created within the target
tumor. This agitates the tissue, which results in frictional heat around the
electrode, and this generated heat is focused and concentrated around the
needle electrode.
0002100651.INDD 423 7/14/2014 8:37:42 PM

Heating of the target tissue at 50°C to 55°C for 4 to 6 minutes produces

irreversible cellular damage. Temperature between 60°C and 100°C pro-
duces irreversible damage to mitochondrial and cytosolic enzymes of the
cells. At more than 100°C, tissue essentially vaporizes. For adequate tumor
tissue destruction, the ablative objective therapy is maintenance of 50°C to
100°C throughout the target tissue volume for 4 to 6 minutes. Slow thermal
conduction through tissues, however, increases the ablation time to 10 to
20 minutes.
In order to achieve low rates of local tumor recurrence, the
interventional radiologist aims to produce a 360-degree, 0.5- to 1.0-cm-
thick tumor-free margin around each tumor. Hence, the diameter of the
ablation must be 1 to 2 cm larger than the diameter of the target tumor.
Incidence of early major complications is 2% to 3% and includes
intraperitoneal bleeding, liver abscess, intestinal perforation, pneumotho-
rax, hemothorax, and bile duct stenosis. Late major complications include
seeding of the needle tract (0.5%). The incidence of minor complications
is 5% to 9% and includes pain, fever, pleural effusion, and asymptomatic
self-limiting intraperitoneal bleeding. The mortality rate is 0.1% to 0.5% and
commonly due to sepsis, hepatic failure, colon perforation, and portal vein
thrombosis.
Indications
1. HCC: A single tumor smaller than 5 cm or as many as three nodules
smaller than 3 cm each without evidence of vascular invasion or extra-
hepatic spread. A good performance status and Child-Pugh class A or B.
2. Secondary liver metastasis: Nonsurgical patients with metastases iso-
lated to the liver. In limited instances, a patient with extrahepatic metas-
tasis may be a candidate for percutaneous treatment, for example, if the
extrahepatic disease is deemed curable.
Contraindications
1. Tumor located less than 1 cm to a main bile duct secondary to increased
risk of stenosis of the central biliary system
2. Intrahepatic biliary dilation
3. Anterior exophytic location of the tumor, due to risk of tumor seeding
4. Untreatable coagulopathy
Relative Contradictions
1. Bilioenteric anastomosis; increased risk of hepatic abscess
2. Superficial lesions; increased rate of complications (e.g., lesions adjacent
to any part of the GI tract as there is an increased risk of thermal injury)
3. Tumors located adjacent to the gallbladder; increased risk of cholecystitis
4. Pacemaker/defibrillators. Can deactivate prior to procedure
Preoperative Workup
Pretreatment imaging must define lesion(s) location relative to surround-
ing structures and organs. Positron emission tomography (PET) may be
indicated to exclude extrahepatic metastatic disease.
Preoperative labs include alpha-fetoprotein (HCC) and carcinoem-
bryonic antigen (colorectal cancer [CRC]), coagulation panel, and CBC.
Antiplatelet therapy should be discontinued for 5 to 10 days prior to
ablation and may be restarted 48 hours postprocedure. Warfarin should be
discontinued 5 days prior and may be restarted 24 hours postprocedure.
002100651.INDD 424 7/14/2014 8:37:42 PM

Postoperative Care
Patients are generally kept at bed rest for 2 hours and admitted for 23-hour
observation.
Repeat imaging evaluation is performed 4 to 8 weeks postproce-
dure (Fig. 34.4). Successful tumor ablation will show nonenhancing abla-
tion region with or without peripheral rim enhancement on CT and MR.
Standard follow-up imaging and evaluation should be performed as recom-
mended for the specific malignancy treated.
FIGURE 34.4 Pre- (A) and post- (B) procedural imaging of a patient undergoing percutane-
ous RFA to treat metastatic colorectal adenocarcinoma (arrow, A). Note a large, nonenhanc-
ing pattern in postablation images (circle, B).
0002100651.INDD 425 7/14/2014 8:37:43 PM

TRANSARTERIAL THERAPY
Over the past 20 years, interventional radiologists have been exploring
novel ways to treat cancer via the transcatheter endovascular approach. In
the last 10 years, these therapies have been studied and refined such that a
new branch of IR has emerged, interventional oncology. Transcatheter ther-
apy provides clinical benefits that are distinct from the traditional medical,
surgical, and radiation oncologic treatments.
Transarterial therapy includes bland embolization, chemoemboliza-
tion, or radioembolization, techniques that selectively treat primary and
metastatic hepatic malignancies via its nutrient arterial supply.
Bland embolization refers to the infusion of embolic particles
via the nutrient artery to cause occlusion of the tumor arterioles.
Chemoembolization refers to selective infusion of chemotherapeutic agents
via the nutrient arterial supply, followed by injection of embolic particles to
increase chemotherapy concentration by preventing chemotherapy wash-
out. An adaptation of chemoembolization with drug-eluting beads (DEBs)
has subsequently been developed. The concept includes loading the che-
motherapeutic agent on a biocompatible, nonresorbable bead that is then
administered via selective catheterization of the tumor’s nutrient arterial
supply. DEBs deliver higher and more sustained chemotherapy doses to the
tumor with reduced systemic exposure.
Radioembolization refers to selective intra-arterial delivery of glass or
resin microspheres loaded with the radioisotope yttrium-90. This delivery
method allows for safe administration of doses that may exceed 150 Gy,
whereas the likelihood of developing severe radiation-induced liver disease
may exceed 50% for external beam radiation doses greater than 40 Gy. In
fact, even higher doses can be attained when a “radiation segmentectomy”
is performed. With this technique, high doses of radiation are delivered to
one or two hepatic segments to maximize tumor irradiation and minimize
exposure of normal liver parenchyma. Calculated segmental radiation doses
in excess of 500 Gy and calculated tumoral doses greater than 1,200 Gy have
been reported, all with very low incidence of biochemical toxicities.
Transarterial Therapy and Hepatocellular Carcinoma

The Barcelona Clinic Liver Cancer (BCLC) staging system for HCC is widely
accepted in clinical practice.
Conventional transarterial chemoembolization (emulsified che-
motherapeutic agent and Lipiodol followed by embolization) is the rec-
ommended first-line therapy in BCLC stage B disease without vascular
invasion, cancer-related symptoms, or extrahepatic spread. Doxorubicin-
loaded DEBs have similar results to conventional chemoembolization in
this population, but with significantly higher administer doses of doxorubi-
cin and significantly reduced serious liver toxicity and doxorubicin-related
adverse events.
An alternative treatment option for HCC patients with BCLC stage B
disease is radioembolization, which has been shown to be safe and effica-
cious specifically in patients with portal vein invasion.
To date, no head-to-head randomized controlled trials have compared
radioembolization versus chemoembolization in HCC patients. However,
a comparative effective analysis of more than 200 HCC patients treated
with radioembolization and chemoembolization suggested similar survival
times, with a significantly reduced toxicity profile for radioembolization.
Transarterial Therapy of Hepatic Metastatic Colorectal Cancer

Both chemoembolization and radioembolization have been used to treat
patients with unresectable metastatic CRC. Radioembolization is used
002100651.INDD 426 7/14/2014 8:37:44 PM

c urrently for the treatment of chemotherapy-resistant or refractory disease.

Treatment of these patients should be considered in the context of the mul-
tidisciplinary group evaluation.
Transarterial Therapy of Neuroendocrine Tumors

Bland embolization, chemoembolization, or radioembolization can be
used in patients with symptomatic but unresectable disease, clinically sig-
nificant tumor burden, or clinically significant progressive disease.
Bland embolization and chemoembolization of neuroendocrine
hepatic metastases provide significant symptomatic improvement and
radiologic responses in the majority of treated patients, with encouraging
progression-free survival.
Radioembolization may provide a complete response in as many as
18% of hepatic neuroendocrine tumor patients, although survival times do
not differ significantly from bland embolization or chemoembolization.
Transarterial Therapy and Intrahepatic Cholangiocarcinoma

Although radioembolization and conventional chemoembolization have
both been shown to be safe and effective in small series of patients with
unresectable cholangiocarcinoma, no randomized control trial has been
performed to date.
Preprocedure Workup
1. Tissue diagnosis or convincing clinical diagnosis
2. CT or MRI of the abdomen and pelvis
3. Exclusion of extrahepatic disease (PET-CT, bone scan, etc.)
4. Lab test: CBC, INR, creatinine, liver function tests, and tumor markers
5. Arteriography shunt study (radioembolization)
Postoperative Care
1. Twenty-three–hour observation in hospital (chemoembolization, DEBs).
2. Outpatient procedure (radioembolization)
3. Aggressive hydration.
4. Symptomatic treatment of postembolization syndrome.
5. Once the patient is tolerating oral intake and pain is controlled, the
patient is discharged to home.
6. Follow-up in IR clinic in 4 to 6 weeks with imaging.
POSTOPERATIVE INTRA-ABDOMINAL FLUID DRAINAGE

Percutaneous drainage of postoperative abdominal fluid collections is one
of the most commonly performed IR procedures. Furthermore, it is a well-
established management option in patients who do not have another indi-
cation for immediate surgery.
Abdominal CT with or without contrast is needed to evaluate for
an appropriate percutaneous window and collection size and to select
the desired approach, that is, transabdominal, transgluteal, and tran-
srectal. Patients with a limited window or small collections can be
approached using CT guidance. More superficial or large collections can
generally be accessed easily under ultrasound and fluoroscopic guid-
ance. The collections are accessed with an 18-gauge trocar needle, and
a stiff guidewire is placed. The drain is then placed through this access,
sutured to the skin, and placed to an external drainage bag. The drain
size depends on the contents of the collection to be drained. To deter-
mine content, a small volume is aspirated through the trocar needle.
Generally, simple pus-filled collections require 8- to 12-Fr drains. A more
0002100651.INDD 427 7/14/2014 8:37:44 PM

complicated collection, such as a peripancreatic collection, may require

14- to 20-Fr drain.
Overall, complication rates are less than 15%, and a 30-day mortal-
ity ranges from 1% to 6% depending largely on the patient’s underlying
medical condition. Major complications include bleeding, septic shock,
enteric fistula, peritonitis, and hemopneumothorax. Minor complica-
tions include pain, infection, pericatheter leak, catheter kinking, and
dislodgement.
Indications
1. Fluid characterization
a. Distinguish purulent fluid, bile, blood, urine, lymph, and pancreatic
secretions.
b. Determine if collection is infected or sterile.
2. Treatment of sepsis
a. Can be curative in patients with simple abscesses
b. Can be curative or temporizing in patients with complex abscesses
3. Relief of symptoms
a. Alleviate pressure and pain due to size or location of collection.
b. Obliterate recurring cysts or collections with sclerosing agents.
Contraindications
Absolute
1. Lack of safe pathway to the collection due to interposed vessels or viscera
Relative
1. Sterile collections: Prolonged catheter drainage may increase risk of sec-
ondary infection.
2. Hematoma: Increased risk of infection and poor drainage.
3. Percutaneous route requires transgression of pleura as this increases the
risk of pneumothorax, empyema, and pleural effusion.
4. Tumor abscess may require lifelong catheter drainage.
5. Echinococcal cyst may elicit anaphylactic reaction if contents leak.
Preoperative Workup
CT scan to evaluate fluid collection for appropriateness. Coagulation stud-
ies. Prophylactic antibiotics (preprocedure antibiotics generally do not
affect cultures)
Postoperative Care (Drain Management)

Strict record of input and output. Flush catheter twice daily with 10 mL
saline to keep catheter free of complex debris. Daily dressing changes. If
patient to be discharged with the drain, proper drain education should be
performed.
If there is high output from the drain (>50 mL/day) beyond 4 days,
the possibility of fistula to the bowel, pancreatic duct, or biliary system
should be considered, and fluoroscopic-guided catheter injection may be
considered.
If fluid collection persists with poor drainage, tissue plasminogen acti-
vator (tPA) or injection drain upsize may be beneficial. Typically, 6 mg of
tPA is diluted in 50 mL of normal saline; up to 50 mL is injected into the
drain twice daily for 3 days. The drain is clamped for 30 minutes after each
injection. This technique can successfully restore drainage in up to 90%
002100651.INDD 428 7/14/2014 8:37:44 PM

of cases, with failure more likely in peripancreatic collections and enteric

fistula.
PSEUDOANEURYSMS
False arterial aneurysms, or pseudoaneurysms (PSAs), are injuries to the
arterial wall, which typically bleed into a saccular hematoma that is con-
tained by the surrounding soft tissues. The name can be a source of confu-
sion: PSAs are not aneurysms, but a manifestation of a vessel injury.
The prevalence of PSAs has increased over time due to increasing
surgical and endovascular procedural volume. Surgery can induce PSA
formation through direct trauma to an artery or through introduction of
infection. Figure 34.5 shows superior mesenteric artery PSA after pancre-
atoduodenectomy. Arterial wall erosion from an adjacent tumor or inflam-
mation (pancreatitis) can cause PSA.
A PSA may be asymptomatic or can cause local symptoms secondary
to mass effect or adjacent tissue ischemia. The most feared complication is
rupture, which can result in life-threatening hemorrhage. PSAs can com-
municate with and rupture into different anatomic spaces, including the
pancreatic duct system, biliary tree, and gut lumen.
Clinically, a PSA can present due to its local or systemic signs and
symptoms (i.e., pain) or can be detected on physical exam with signs
such as a pulsatile mass, an audible bruit, or a palpable thrill. They can
FIGURE 34.5 Angiography demonstrating large superior mesenteric

artery pseudoaneurysm (arrow) sustained after pancreatoduodenectomy.
0002100651.INDD 429 7/14/2014 8:37:45 PM

otentially manifest with a “sentinel bleed,” such as bleeding into a drain or

p
as a GI bleed (e.g., hematemesis, melena). The diagnosis of PSA is secured
with imaging.
PSAs can be evaluated with computed tomography angiography (CTA)
and magnetic resonance angiography (MRA). Both modalities can demon-
strate deeper arterial structures and can be used for planning intervention.
Contrast-enhanced CTA or MRA can demonstrate a contrast-opacified,
saccular structure communicating with a blood vessel, with unopacified
areas in the sac representing thrombus. CTA is fast and not operator depen-
dent; however, it requires the use of iodinated contrast media, which can be
nephrotoxic, as well as ionizing radiation. MRA lacks ionizing radiation but
is time consuming and limited by motion and other artifacts, which makes
it a suboptimal modality for PSA detection.
Another modality for imaging of PSAs is percutaneous, catheter-
directed arteriography, which involves real-time, dynamic imaging of arte-
rial flow and also allows definitive endoluminal therapeutic intervention.
Catheter-directed arteriography however is invasive and also involves
ionizing radiation and iodinated contrast media. Potential complications,
while rare, include arterial vessel damage, including new PSA formation,
dissection, rupture, thrombosis, and hematoma.
Management of PSAs has classically been surgical; however, with
advances in image-guided percutaneous management, today, virtually all
visceral PSAs are successfully managed with IR techniques.
The purpose of endoluminal management is PSA exclusion from the
circulation. The specific endoluminal treatment depends on whether or not
the blood vessel supplying the PSA is expendable or not and whether or
not collateral circulation exists. If the vessel is not expendable, a covered
stent can be placed across the neck of the PSA. If the vessel is expendable
and is an end artery without collateral supply, the artery feeding the PSA
can be coil embolized. If the vessel is expendable but has collateral circula-
tion, the PSA must be “bracket” embolized, with coils placed in the artery
central and peripheral to the PSA. Complications of endoluminal manage-
ment include “nontarget” embolization, or coils migrating to unintended
locations, as well as blood vessel rupture. Embolized PSAs and feeding
vessels can also occasionally recanalize. Stent graft placement is relatively
contraindicated with mycotic aneurysms due to risk of stent graft infec-
tion. Larger arteries have a lower risk of in-stent thrombosis than smaller
arteries. A small risk of rebleeding exists after IR treatment of visceral PSA;
patients should be counseled accordingly.
CONCLUSION
A diverse array of IR techniques supports HPB surgery today. Close collabo-
ration between HPB surgeon and interventional radiologist is necessary to
optimize diagnosis and treatment of complex HPB pathology.
Suggested Readings
Boyer TD, Haskal ZJ. AASLD practice guidelines: the role of transjugular intrahe-
patic portosystemic shunt (TIPS) in the management of portal hypertension.
Hepatology 2010;51:306.
Cirocchi R, Trastulli S, Boselli C, et al. Radiofrequency ablation in the treatment of liver
metastases from colorectal cancer. Cochrane Database Syst Rev 2012;6:CD006314.
doi: 10/1002/14651858.CD006317.pub3
Mahnken AH, Pereira PL, de Baere T. Interventional oncologic approaches to liver
metastases. Radiology 2013;266(2):407–430. doi 10.1148/radiol.12112544. PMID:
23362094 (Pub Med-indexed for MEDLINE)
002100651.INDD 430 7/14/2014 8:37:45 PM

Pitt HA, Venbrux AN, Coleman J, et al. Intrahepatic stones: the transhepatic team
approach. Ann Surg 1994;216(5):527–535; discussion 535–537. PMID: 8185402(Pub
Med-indexed for MEDLINE)
van Lienden KP, et al. Portal vein embolization before liver resection: a systematic
review. Cardiovasc Intervent Radiol 2013;36:25–34.
Zyromski NJ, Viera C, Stecker M, et al. Improved outcomes in postoperative and pan-
creatitis-related visceral pseudoaneurysms. J Gastrointest Surg 2007;11(1):50–55.
0002100651.INDD 431 7/14/2014 8:37:45 PM

002100651.INDD 432 7/14/2014 8:37:45 PM
INDEX
Note: Page numbers followed by “f ” indicate figures; those followed by “t” indicate
tabular material.
A outcomes of, 25
Abdominal pain pathophysiology of, 14
in acute pancreatitis, 15 Acute variceal bleeding, 134–137
in chronic pancreatitis, 28 balloon tamponade, 135
in SOD, 302 endoscopic treatment, 135
Abdominal x-rays for gallbladder disease, medications for, 134–135
286 octreotide infusion, 135
Ablation technique surgical shunt creation, 136–137
for colorectal liver metastases, 183–184 TIPS procedure, 136
for hepatocellular carcinoma, 238–239 treatment options for, 135, 136f
for metastatic colorectal cancer, 241 vasopressin administration, 135
for neuroendocrine carcinoma, 198 Adjuvant chemotherapy, for PDAC, 49
Ablative technologies Adjuvant therapy, for PDAC, 48–49
microwave energy technology, 163 Aloka ultrasound unit, console of, 140, 141f
radiofrequency-assisted parenchymal Alveolar echinococcosis (AE), 207, 210–211
transection, 162–163, 166 Amebic liver abscess
Accessory duct of Santorini, 3, 4 diagnosis of, 213
Acinar cells, 11 presentation of, 212
carcinoma of, 39 treatment of, 213
Acute ascending cholangitis. See American Association for the Study of Liver
Cholangitis Diseases (AASLD), for HCC, 169, 170f
Acute cholecystitis, 290. See also American Joint Committee on Cancer
Cholecystitis (AJCC) staging
cholecystectomy for, 291 for hilar bile duct cancer, 335, 337t
cholecystoenteric fistula, 293–296 for pancreatic neuroendocrine
emphysematous cholecystitis, 292, 293f neoplasms, 73, 74t
gallbladder for PDAC, 43, 44t
empyema, 291, 292f Angiomyolipomas (AML)
gangrene of, 292 diagnosis of, 218
perforation, 292, 294f presentation of, 218
outcomes of, 298 treatment of, 220
Acute pancreatitis (AP) Annular pancreas (AP), 4–5
diagnosis of, 14–18 Antibiotics
epigastric pain, 14 for acute cholangitis, 276t
features, 14–15 for necrotizing pancreatitis, 22
imaging in, 16–18 for SAP management, 20
severity of, 15–16 AP. See Acute pancreatitis (AP)
endoscopic retrograde Argon beam coagulator, 162
cholangiopancreatography, 402 Arterial blood supply, of pancreas, 5–7, 6f
etiology of, 14, 15t Asymptomatic gallstones, 283
incidence of, 14 Ataxia-telangiectasia, 40
management of, 18–25 Atlanta classification, for AP severity, 16
mild AP, 18–19
necrotizing pancreatitis, 19, 21–23 B
severe acute pancreatitis, 19–25 Bacteriology, in biliary surgery, 271–272
onset of, 15 Balloon dilation, for SOD treatment, 308
433
0002179469.INDD 433 7/14/2014 7:58:24 AM

434 Index
Balloon tamponade, for acute variceal extrahepatic, 268–269

bleeding, 135 and liver resections, 259–260, 261f
Band ligation, for acute variceal bleeding, gallbladder
135 agenesis of, 266
Barcelona Clinic Liver Cancer (BCLC) cystic artery, 266–267
staging system, for HCC, 171, 171f, cystic duct, 266, 267f
234, 235, 426 double gallbladder, 266
Beger procedure, for chronic hepatocystic triangle, 266
pancreatitis, 33, 35f–36f liver
Benign cystic liver lesions and bile duct drainage, 259–260, 261f
amebic liver abscesses, 212–213, 214t caudate lobectomy, 255
cystadenoma, 206–207 fibrous capsules and attachments,
hydatid cysts, 207–208, 208f, 209t, 265–266
210–211 hemihepatectomies, 258f
polycystic liver disease, 205–206, 206f hepatic artery (See Hepatic artery)
pyogenic liver abscesses, 211–212, 212f and hepatic veins, 262–263, 263f
simple hepatic cysts, 204–205 plate-sheath system, 263–265, 264f,
Benign solid liver lesions 265f
angiomyolipomas, 218, 220 and portal veins, 260, 262, 262f
focal nodular hyperplasia, 215 prevailing pattern, 260f
hemangiomas, 213, 215 sectionectomies, 258f
hepatic adenomatosis, 217 segmentectomies, 259f
hepatocellular adenoma, 215–217, 217f trisectionectomies, 259f
management of, 219t, 220 Biliary cysts, classification system of, 325,
nodular regenerative hyperplasia, 326f, 327
217–218 Biliary ducts
Bile duct injury (BDI) choledochal cysts (See Choledochal
etiology of, 357–359 cysts (CCs))
incidence of, 354 computed tomography, 392–393
management of, 365–370 endoscopic retrograde
algorithm for, 365, 366f cholangiopancreatography, 394
intraoperative cholangiogram, magnetic resonance
366, 367f cholangiopancreatography,
postoperative cholangiogram, 369f 393–394, 394f
primary end-to-end repair, 367–368 percutaneous transhepatic
Roux-en-Y hepaticojejunostomy, cholangiography, 394
367, 368f ultrasound, 392
Roux-en-Y limb of jejunum, 368–369 Biliary dyskinesia, 286
outcomes/follow-up of, 370 Biliary fistula, 107. See also Bile leak
Bile ducts minimally invasive biliary surgery,
blood supply of, 269–270, 269f 351, 352
complications of LT Biliary leak
anastomotic strictures, 378 clinical presentation of, 359
biliary leak, 378 complications of LT, 378
intrahepatic strictures, 378 definition of, 354
extrahepatic, 268–269 diagnosis of, 359–361
and liver resections, 259–260, 261f cholangiography, 359
Bile leak endoscopic retrograde
CT scan, 227 cholangiopancreatography,
definition of, 226 359–360, 361f
diagnosis and management of, HIDA scintigraphy, 360
226–227, 228f laboratory studies, 359
endoscopic retrograde magnetic resonance
cholangiopancreatogram, 227 cholangiopancreatography, 360, 363f
percutaneous transhepatic percutaneous transhepatic puncture,
cholangiogram, 227 359, 360, 362f
preoperative and intraoperative ultrasonography, 359
considerations, 227–228 etiology of, 354–357, 355t
Bile salts, 272t, 273 mangement of, 361–365
Biliary anastomosis, 82 after cholecystectomy, 361–362
Biliary anatomy after liver resection, 362, 364
bile ducts after liver transplantation, 364
blood supply of, 269–270, 269f after trauma, 364–365
002179469.INDD 434 7/14/2014 7:58:24 AM

Index 435
Biliary malignancies Cava-preserving technique, for LT, 375,

diagnosis of, 333–335 377f, 378
etiology of, 332–333 Cavitron Ultrasonic Surgical Aspirator
management of (CUSA), 163–164
preoperative planning, 336–340 CBD stones. See Common bile duct
preoperative staging, 335–336 (CBD) stones
operative indications and therapy, Celiac plexus block (CPB), 405–406, 406f
340–342 Celiac plexus neurolysis (CPN), 405–406
outcomes of, 342–343, 342f Central pancreatectomy, 87
postoperative complications, Central venous pressure (CVP)
prevention and treatment of, 342 anesthesia concept, for liver
Biliary obstruction, 274 resections, 225
Biliary sludge, 279 Centroacinar cell, 11
Biliary stenting, for SOD treatment, 309 Chemoembolization, 237, 426, 427
Biliary surgery Chemotherapy
infections in, 274–275 for colorectal liver metastases, 183, 183f
acute ascending cholangitis, for hepatocellular carcinoma, 239
275–277, 276t for metastatic colorectal cancer, 240
acute cholecystitis, 277–278 for neuroendocrine carcinoma, 199–201
bacteriology, 271–272 Chemotherapy-associated liver injury
etiology of obstruction, 274 (CALI), 230
hepatic resection, 274, 275 Child-Pugh score, 113, 114t, 234
host defense mechanism, 272–273, Child-Turcotte-Pugh (CTP) score, 130,
272t 130t, 138
inferior and superior mesenteric Cholangiocarcinoma (CCA)
venous systems, 271 definition of, 239
pathophysiology, 274 liver transplant for, 240
minimally invasive (See Minimally yttrium-90 microspheres, 240
invasive biliary surgery) Cholangitis
Biliary system, IOUS of, 148–149, causes of, 275
149f, 150f diagnosis of, 275, 276
Biliary-enteric anastomosis, 350–351, 350f drainage of, 276, 277
Bilomas. See Bile leak Reynolds pentad, 275
Biochemical markers, of AP severity, Cholecystectomy, 19
15, 16t for acute cholecystitis, 291
Bipolar electrosurgery, 159 biliary leak mangement after, 361–362
Bipolar vessel-sealing devices, 167t Cholecystitis
Enseal, 160–161 acute, 284–286, 285f
LigaSure, 159–160 chronic, 284
PlasmaKinetic tissue management infections in biliary surgery
system, 161–162 causes of, 277
salient dissecting sealer, 161 complications, 277, 278
Bismuth-Corlette classification scheme diagnosis of, 277
for hilar cholangiocarcinoma, 341 gallbladder perforation, 278
of malignant biliary stricture, 335, 336f laparoscopic cholecystectomy, 277,
Bladder drainage technique, 381, 382 278
Bland embolization, 426 Cholecystokinin-Tc-HIDA scan, for
for HCC, 237 biliary dyskinesia, 286
Borderline resectable tumor, 43–44 Choledochal cysts (CCs)
Botulinum toxin (Botox) injection, for classification system, 325, 326f, 327
SOD, 309 definition of, 325
diagnosis
C imaging techniques, 327–328
Calcineurin inhibitors, 380 presentation of, 327
Cambridge Classification System, for etiology of, 325
chronic pancreatitis, 29t outcome/follow-up, 330
Cancer of the Liver Italian Program prevalence of, 325
(CLIP) system, 234, 235 surgical management
Carcinoembryonic antigen (CEA), 57, 76, complete cyst resection, 328, 329
180, 334 intrahepatic disease and cirrhosis,
Carcinoid syndrome, 188, 188f 329, 330
Caroli disease, 326f, 327, 329, 330 intrahepatic involvement, 329
Caudate lobectomy, 255 simple cyst excision, 329
0002179469.INDD 435 7/14/2014 7:58:24 AM

436 Index
Choledochocele, 327 Classic laparoscopic BDI, 357, 358f

Choledochoduodenostomy, for CBD Classical vs. pylorus-preserving PD, 47
stones, 323 Coagulopathy, laparoscopic resection for
Choledocholithiasis, 393f HCC, 246
Cholescintigraphy Colorectal cancer (CRC), risk factors
gallbladder, 391–392, 392f for, 178
gallbladder disease, 287–288, 288f Colorectal liver metastases (CRLM)
Cholestatic liver disease, 373, 374t diagnosis of, 178–180
Chronic cholecystitis management of
laparoscopic cholecystectomy for, operative strategy, 184
288–290, 290f preoperative considerations,
outcomes of, 296, 298 180–184
Chronic pancreatitis (CP) outcomes/follow-up, 185, 185t
clinical presentation of, 28 resectability for, 180–181, 181t
diagnosis of, 29 Colorectal metastases, laparoscopic liver
endoscopic management of, 29–30 resection for, 245–246
endoscopic retrograde Combined endoscopic and percutaneous
cholangiopancreatography, approach, for SAP management, 23
401–402 Common bile duct (CBD) stones
epidemiology of, 27 classification of, 314
features of, 151, 152f diagnosis of
genes implicated in, 28t direct cholangiography, 316–317
pathogenesis of, 27–28 elevated liver profile values, 315
risk factors for, 27 endoscopic ultrasound, 316
surgical management of, 30–37 intraoperative cholangiography,
Beger procedure, 33, 35f–36f 317, 317f
distal pancreatectomy, 35 intraoperative ultrasonography, 317
duodenum-preserving pancreatic magnetic resonance
head resection, 33, 35f–36f, 37t cholangiopancreatography,
Frey procedure, 31, 34f 315–316, 315f
goals for surgery, 30 serum liver function tests, 314
LDP, 37 transabdominal ultrasonography,
longitudinal pancreaticojejunostomy, 315
31, 32t–33t drainage procedures for, 323
minimally invasive pancreatic endoscopic therapy for, 318–319
surgery, 37 etiology of, 314
pancreatoduodenectomy, 31 incidence of, 314
TP-IAT, 36–37 laparoscopic choledochotomy,
ultrasound characteristics of, 152t 321–322, 321f, 322t
Cirrhosis laparoscopic therapy for, 319–320
acute variceal bleeding (See Acute LTCBDE
variceal bleeding) indications for, 319–320
Child-Pugh score, 113, 114t vs. laparoscopic choledochotomy,
clinical manifestations of 319t
ascites, 132–133 technique of, 320–321, 320f
esophageal varices, 134 management, algorithm for, 318, 318f
hepatic encephalopathy, 132 open CBD exploration, 322
pulmonary complications, 131–132 secondary, 314
renal, 130–131 sonographic criteria for, 148, 150f
spontaneous bacterial peritonitis, transduodenal sphincteroplasty for,
134 323
diagnosis of, 129–130 Complete cyst resection procedure, for
etiology of choledochal cysts, 328, 329
causes of, 127, 128t, 129 Computed tomography (CT)
pathophysiology of, 126–127 for acute pancreatitis, 17–18
evolution of, 126, 127f bile leak, 227
liver transplantation, 137–138 biliary ductal system, 392–393
MELD, 113 choledochal cysts, 328
physical findings in, 129t for chronic pancreatitis, 29
preoperative considerations of, 138 CRLM diagnosis, 178–179
stellate cells, 126 cystic echinococcosis, 207, 208f
Cirrhosis, laparoscopic resection for gallbladder, 391
HCC, 246 gallstone disease, 288, 289f
002179469.INDD 436 7/14/2014 7:58:24 AM

Index 437
hemangiomas, 213 Direct cholangiography, for CBD stones,

hepatocellular adenoma, 216, 217f 316–317
liver, 388 Dissection-only devices
macrocystic SCA, 51, 52f CUSA, 163–164
MCN with associated mural water-jet dissector, 164
calcification, 55f Distal pancreatectomy (DP)
neuroendocrine carcinoma, 191f, 192f, in chronic pancreatitis, 35
193f description, 83–84
pancreas, 394 hybrid approach, 86
pancreatic ductal system, 392–393 laparoscopic approach, 86
pancreatic lymphoepithelial cyst, 51, lateral to medial approach, 84–85
52f medial to lateral approach, 85
pancreatic neuroendocrine neoplasms, outcomes of, 94
67, 68f pancreatic adenocarcinoma, 46–47
for PDAC, 42 pancreatic fistula incidence, 100
of pyogenic liver abscess, 211, 212f for PDAC resection, 46–47
Computed tomography cholangiography splenic preservation, 85–86
(CTCP), choledochal cysts, 328 DNA mismatch repair genes, for
Conventional approach, for LT. See Cava- PDAC, 41
preserving technique, for LT Doxorubicin-eluting beads, 237
Critical view of safety (CVS), 357, 358f, Drug-eluting beads (DEBs)
359 hepatocellular carcinoma, 237
CRLM. See Colorectal liver metastases metastatic colorectal cancer, 241
(CRLM) Ducts of Luschka, 267–268
CUSA. See Cavitron Ultrasonic Surgical Duplex and color flow Doppler imaging,
Aspirator (CUSA) 387
Cyst fluid analysis, 57–58
Cystic artery, gallbladder anatomy, E
266–267 Echinoccocus granulosus, 207. See also
Cystic duct, gallbladder anatomy, 266, Cystic echinococcosis (CE)
267f Echinoccocus multilocularis, 207, 210.
Cystic echinococcosis (CE) See also Alveolar echinococcosis
cystectomy of, 210 (AE)
presentation of, 207 Electrosurgery
WHO classification of, 207, 209t bipolar, 159
Cystic lesions, benign. See Benign cystic complications of, 158–159
liver lesions limitation of, 158
Cystic lesions, of pancreas. See Pancreatic principle of, 157
cystic lesions Embolization, HCC
Cystic pancreatic neuroendocrine bland particle embolization, 237
tumors (pNETs), 53 chemoembolization/drug-eluting
Cystic plate beads, 237
gallbladder anatomy, 267–268 locoregional therapies, 175–176
plate-sheath system of liver, 265 radioembolization, 237–238
Cysts Endocrine graft drainage technique, 382
biliary cysts, classification system of, Endocrine insufficiency, incidence of,
325, 326f, 327 109
choledochal cysts (See Choledochal Endocrine pancreas
cysts (CCs)) cells of, 11, 12t
Cytotoxic chemotherapy, 72 function of, 11
Endoscopic band ligation, for acute
D variceal bleeding, 135
Daptomycin, 275 Endoscopic necrosectomy, for SAP
Delayed gastric emptying (DGE), 48 management, 23
definition of, 103 Endoscopic retrograde
etiology of, 103–104 cholangiopancreatography (ERCP),
incidence of, 103 397
treatment acute pancreatitis, 402
NG tube placement, 104, 105 bile leak, 227
peripherally inserted central biliary ductal system, 394
catheter, 104–105 CBD stones, 316f
promotility agent, 104 chronic pancreatitis, 401–402
Diagnostic laparoscopy, in PDAC, 44–45 IPMN, 56–57
0002179469.INDD 437 7/14/2014 7:58:24 AM

438 Index
Endoscopic retrograde LigaSure, 159–160, 166

cholangiopancreatography (ERCP) PlasmaKinetic tissue management
(Continued) system, 161–162
jaundice salient dissecting sealer, 161
biliary obstruction and drainage, 400 dissection-only devices
gastric outlet obstruction, 401 CUSA, 163–164
self-expandable metallic stents, water-jet dissector, 164
400, 401 electrosurgery (See Electrosurgery)
pancreatic ductal system, 394 finger fracture technique, 157
pancreatic fistula treatment, 101 ultrasonic cutting and coagulating
PDAC, 43 surgical device, 164–165
in periampullary malignancy Enseal, 160–161, 165, 166, 168
distal bile duct stricture, 399, 399f Entamoeba histolytica infections, 212. See
tissue sampling, 400 also Amebic liver abscess
postoperative anatomy, 406–407 Enteric exocrine drainage technique, 381
postoperative complications, 404 Enzyme-linked immunoabsorbent assay
preprocedure evaluation, 398–399, 399f (ELISA), 213
Roux limbs, 407 Escherichia coli, 211
Endoscopic retrograde pancreatography Everolimus, 72, 73, 201
(ERP), for chronic pancreatitis, 29 Exocrine insufficiency, incidence of, 107
Endoscopic sphincterotomy Exocrine pancreas
for CBD stones, 318–319 functional unit of, 11
for SOD, 306–308, 307t solid epithelial neoplasms of, 39
Endoscopic ultrasound (EUS) Extended lymphadenectomy, for
for acute pancreatitis, 17 PDAC, 47
for CBD stones, 316 Extracorporeal shock wave lithotripsy
for chronic pancreatitis, 29, 30t (ESWL), 402
fine needle aspiration, 397, 405 Extrahepatic bile ducts, 268
mural nodularity within branch-type Extrahepatic biliary tract injuries, 412
IPMN, 56, 57f
palliation of pain, 405–406, 406f F
pancreatic neuroendocrine neoplasms, False arterial aneurysms. See
67, 69f, 70 Pseudoaneurysms (PSAs)
for pancreatobiliary drainage, 405 Familial atypical multiple mole
for PDAC, 43 melanoma (FAMMM) syndrome, 40
preprocedure evaluation, 398–399, 399f Fibrolamellar hepatocellular carcinoma,
staging of, 405 215
for tissue sampling, 404–405 Fibronectin, 273
Endoscopy Fibrous capsules, liver, 265–266
ERCP (See Endoscopic retrograde Fine needle aspiration (FNA), 204
cholangiopancreatography (ERCP)) EUS, 397, 405
EUS for SAP management, 21
fine needle aspiration, 397, 405 Finger fracture technique, 157
palliation of pain, 405–406, 406f Fistula
for pancreatobiliary drainage, 405 biliary, 107
preprocedure evaluation, 398–399, complications of pancreas
399f transplantation, 383, 384
staging of, 405 pancreatic, 99–103
for tissue sampling, 404–405 Fluid resuscitation, of SAP, 20
pancreatic fluid collections Focal nodular hyperplasia (FNH), 148, 215
pseudocyst, 402, 403 Frey procedure, 31, 34f
transmural drainage, 404 Future liver remnant (FLR), 180–182
transpapillary drainage, 403, 403f assessment and optimization, 117
End-to-end repair, 367, 368 hepatectomy, 113, 115f
Energy devices, for parenchymal postoperative hepatic insufficiency,
transection 228–230
ablative technologies
microwave energy technology, 163 G
radiofrequency-assisted Gallbladder
parenchymal transection, 162–163 adenocarcinoma, survival for, 342, 343f
argon beam coagulator, 162 anatomy of
bipolar vessel-sealing devices, 167t agenesis of, 266
Enseal, 160–161, 165, 166, 168 cystic artery, 266–267
002179469.INDD 438 7/14/2014 7:58:24 AM

Index 439
cystic duct, 266, 267f and liver resections, 257–259, 260f

double gallbladder, 266 ramification of
hepatocystic triangle, 266 first-order division, 254, 255f, 256f
cholescintigraphy, 391–392, 392f second-order branches, 254, 255f,
computed tomography, 391 256f
ultrasound, 390–391, 391f third-order branches, 254, 255, 255f,
Gallbladder disease 257f
diagnostic imaging in Hepatic injuries
abdominal x-rays, 286 follow-ups, 411
cholescintigraphy, 287–288, 288f management
computerized tomography, 288, 289f complications, 411
magnetic resonance imaging, 288, operative care, 410–411
289f postoperative management, 411
ultrasound, 286–287, 287f preoperative considerations,
etiology/pathogenesis of 409–410
cholecystitis, 284–286 technical tips, 410–411
gallstones, 279–284 outcomes, 411
management of Hepatic metastatic colorectal cancer,
acute cholecystitis, 290–291 transarterial therapy of, 426–427
chronic cholecystitis, 288–290, 290f Hepatic neuroendocrine metastases, 72
severe cholecystitis, 291–296 Hepatic resection, 274
outcomes of, 296, 298 hepatocellular carcinoma, 216
prevalance of, 279 NEC, 188
Gallstone ileus, 295f Hepatic tight junctions, 272
management of, 295–296 Hepatic veins, and liver resections,
occurrence of, 293 262–263, 263f
signs and symptoms of, 294 Hepatitis B, 236
Gallstones Hepatobiliary host defense mechanism,
biliary sludge, 279 272, 272t, 273
cholesterol, 279–282, 280f Hepatocellular adenoma (HCA)
natural history of, 283–284, 283t beta-catenin gene mutations, 216
pigment, 281f, 282 classification of, 216
prevalence of, 282 diagnosis of, 216
risk factors for, 282 presentation of, 215–216
Gastric outlet obstruction, treatment risk factors for, 215
of, 401 treatment for, 216–217, 217f
Gastrinoma triangle, 65, 66f Hepatocellular carcinoma (HCC)
Gastrinomas AASLD practice guidelines, 170f
in MEN1, 71 ablation, 238–239
pancreatic neuroendocrine neoplasms, characteristics, 170
65, 66f chemotherapy, 239
clinical staging systems, 234, 235
H diagnosis of, 170, 234
Hand-assisted technique, 245, 250 embolization
Harmonic scalpel, 164–166 bland particle embolization, 237
HCA. See Hepatocellular adenoma (HCA) chemoembolization/drug-eluting
Heat shock protein-32, 273 beads, 237
Heat-sink effect, ablative technique, 239 radioembolization, 237–238
Hemangiomas factors, 233
diagnosis of, 213 guidelines, 234
presentation of, 213 laparoscopic resection for, 246–247
symptoms of, 213 liver transplantation, 235, 373, 374, 374t
treatment of, 213, 215 locoregional therapies, 174–176
Hepatectomy ablation, 175
complications (See Liver surgery, embolization, 175–176
complications of) limitations and outcomes, 175
future liver remnant, 113, 115f selection criteria, 176
laparoscopic liver surgery, 245 magnetic resonance imaging,
liver transplantation, 375, 376f, 377f, 389f–390f
378 risk factors, 169
Hepatic adenomatosis, 217 staging systems for, 171, 171f
Hepatic artery stereotactic body radiation therapy, 238
anomalies of, 268 surgical resection, 235–236
0002179469.INDD 439 7/14/2014 7:58:24 AM

440 Index
Hepatocellular carcinoma (HCC) postoperative intra-abdominal fluid

(Continued) drainage, 427–429
surveillance of, 169–170 preoperative portal vein embolization,
and transarterial therapy, 426 422–423
treatment of pseudoaneurysms, 429–430
liver resection, 172–173 radiofrequency ablation, 423–425, 425f
orthotopic liver transplant, 173–174 transarterial therapy, 426–427
transplantation and resection, 171 transjugular intrahepatic
Hepatofugal flow, 136 portosystemic shunt, 419–422
Hepatopulmonary syndrome (HPS), Intra-abdominal fluid drainage. See
131, 374 Postoperative intra-abdominal
Hepatorenal syndrome (HRS), 130–131, fluid drainage
131t, 374 Intraductal papillary mucinous
Hereditary breast and ovarian neoplasm (IPMN), 53
cancer, 40 cancer antigen (CA) 19-9, 54
Hereditary nonpolyposis colorectal CEA level, 57
cancer (HNPCC type II), 40 cytopathology, 57
Hereditary pancreatitis, 40 degree of oncologic risk, 59
Hilar bile duct cancer, AJCC staging ductal connection and multifocality,
system, 335, 337t 55
Hilar CCA, liver transplant for, 239 ERCP, 56–57
Hilar cholangiocarcinoma EUS, 56, 57f
MRCP for, 336–337, 339f MRCP, 55
preoperative staging system for, 335, MRI with MRCP, 395, 395f
338t mucin-laden papillary epithelial
Hilar plate, 264 architecture, 53, 53f
HO-1 and HO-2 enzymes, 273 PET, 56
Hormonal therapy, for neuroendocrine preoperative considerations, 59
carcinoma, 199 Intrahepatic cholangiocarcinoma
Host defense mechanism diagnosis of, 334
bile salts, 273 incidence of, 333
hepatic tight junctions, 272, 273 pathologic features of, 343
hepatobiliary, 272, 272t, 273 phenotypes of, 341
immunoglobulin A, 273 and transarterial therapy, 427
Kupffer cells, 273 Intraoperative cholangiography (IOC)
Hydatid (echinococcal) cyst in BDI prevention, 359
alveolar echinococcosis, 210–211 for CBD stones, 317, 317f
cystic echinococcosis, 207–208, 208f, 210 Intraoperative hemorrhage
definition of, 222, 223t, 224
I diagnosis and management of, 224
Immunoglobulin A (IgA), 273, 274 preoperative and intraoperative
Immunosuppression, 383 considerations, 225–226
Indocyanine green retention rate at 15 Intraoperative laparoscopic ultrasound,
minutes (ICG R15), 115 155–156, 156f
Infectious cholangitides. See Biliary surgery Intraoperative ultrasound (IOUS)
Inferior vena cava (IVC) ligament basic physics of, 140
hepatic veins, 263 of biliary system, 148–149, 149f, 150f
right hepatectomy, 250 for CBD stones, 317
Inflammatory HCA, 216 for colorectal liver metastases, 184
Insulinomas, 65 of liver, 148, 388
International Hepatopancreatobiliary guidance, 145
Association (IHPBA) Consensus systematic approach, 146–147
Brisbane (2000) classification for mechanics of, 140–145
liver surgery, 111, 114f Aloka ultrasound unit, console of,
International Study Group on Pancreatic 140, 141f
Fistula (ISGPF), 99 annotation and storage, 145
Interventional radiology (IR) technique, biopsy/ablation, 145
for HPB patients depth, 143
percutaneous biliary intervention (See Doppler mode, 145
Percutaneous biliary intervention, echogenicity, 141
in biliary obstruction) image refinement, 143
percutaneous transhepatic orientation, 142, 143f
cholangiography, 415, 417f probe types, 142
002179469.INDD 440 7/14/2014 7:58:24 AM

Index 441
scanning techniques, 143, 144f hepatic hilar dissection, 92–93

signal transmission, 142 outcomes of, 94
of pancreas, 151–154 pancreatic head resection, 93
pancreatic duct, 152 pancreatic neck dissection, 91–92
pancreatic parenchyma, 151, 152f specimen removal, 93
target lesion, evaluation of, 153–154, trocar positioning for, 91, 92f
153f, 154f, 155f Laparoscopic reconstruction
vascular relationships, 153 procedure vs. open biliary-enteric
scanning approach, 145t reconstruction, 351t
sound waves, 140 Laparoscopic transabdominal
IOUS. See Intraoperative ultrasound debridement, for SAP
(IOUS) management, 24
Irinotecan-eluting beads, 241 Laparoscopic transcystic duct bile duct
Ischemia, of colon/gallbladder, 21 exploration (LTCBDE)
indications for, 319–320
J vs. laparoscopic choledochotomy, 319t
Jaundice, ERCP technique of, 320–321, 320f
biliary obstruction and drainage, 400 LDP. See Laparoscopic distal
gastric outlet obstruction, 401 pancreatectomy (LDP)
self-expandable metallic stents, Left lateral sectionectomy
400, 401 laparoscopic liver surgery, 247, 249,
Jejunojejunostomy leak, 352 250f
LigaSure, 159–160, 166
Liver
K
anatomic resections
Klebsiella pneumoniae, 211
caudate lobectomy, 255
Kupffer cells, 273, 274
hemihepatectomies, 258f
prevailing pattern, 260f
L sectionectomies, 258f
β-Lactamase inhibitor, 275 segmentectomies, 259f
Lanreotide, 72 trisectionectomies, 259f
Laparoscopic approach, to pancreatic anatomy, 111, 112f, 113f
malignancy, 97 benign cystic lesions (See Benign cystic
Laparoscopic cholecystectomy, for liver lesions)
chronic cholecystitis management, benign solid lesions of
288–290, 290f angiomyolipomas, 218, 220
Laparoscopic distal pancreatectomy focal nodular hyperplasia, 215
(LDP) hemangiomas, 213, 215
for chronic pancreatitis, 37 hepatic adenomatosis, 217
outcomes of, 94 hepatocellular adenoma, 215–217,
pancreatic mobilization, 90–91 217f
pancreatic resection, 91 nodular regenerative hyperplasia,
pancreatic stump management and 217–218
closure, 91 and bile duct drainage, 259–260, 261f
specimen retrieval, 91 cancer treatment platforms (See
trocar positioning for, 89, 90f Multidisciplinary liver tumor
Laparoscopic liver surgery group)
anatomic resections, 247–251 computed tomography, 388
left lateral sectionectomy, 247, 249, duplex and color flow Doppler imaging,
250f 387
right hepatectomy, 247–251, 251f fibrous capsules and attachments,
hand-assisted technique, 245 265–266
hepatectomy, 245 hepatic artery (See Hepatic artery)
limitations, 243 and hepatic veins, 262–263, 263f
oncologic applications injury characteristics, chemotherapy-
colorectal metastases, 245–246 associated, 117, 117t
hepatocellular carcinoma, 246–247 intraoperative ultrasound, 148, 388
recommendations for, 244 guidance, 145
robotic surgery, 248–249 systematic approach, 146–147
Laparoscopic pancreaticoduodenectomy, magnetic resonance imaging, 388,
95t 389f–390f, 390
anastomotic reconstruction, 93 physiology, 112–115
duodenal mobilization, 92 plate-sheath system, 263–265, 264f, 265f
0002179469.INDD 441 7/14/2014 7:58:24 AM

442 Index
Liver (Continued) procedure/technique

portal vein anatomy, 111, 112f graft implantation, 378, 379f
and portal veins, 260, 262, 262f organ procurement operation, 375,
positron emission tomography 376f
scan, 390 preparation of graft, 375
preoperative evaluation recipient hepatectomy, 375, 377f, 378
complications, risk factors, 116 Liver-directed therapies (LDTs)
initial assessment, 116–119 for HCC, 235, 374
preoperative therapy and downsizing, neuroendocrine carcinoma, 187
119–123 Liver-related mortality
resection definition of, 228–229
biliary leak mangement after, 362, diagnosis and management of, 229
364 preoperative and intraoperative
characteristics, 172 considerations, 229–230
evaluation, future liver remnant Local resection of pancreatic
volume, 172–173 head with longitudinal
outcomes, 173 pancreaticojejunostomy (LR-LPJ),
portal vein embolization, 173 31, 34f
preoperative assessment, liver Locally advanced tumor, 44
function, 172 Long common channel theory, 325
sonography, 387 Low CVP anesthesia concept, for liver
ultrasound, 387–388 resections, 225
Liver surgery, complications of LT. See Liver transplantation (LT)
bile leak and organ space infections, Lymphatic drainage, of pancreas, 9, 10f
226–228 Lymphoepithelial cysts, 51, 52f
definition of, 226
diagnosis and management of, M
226–227, 228f Magnetic resonance
preoperative and intraoperative cholangiopancreatography (MRCP)
considerations, 227–228 biliary ductal system, 393–394, 394f
hepatectomies, 221 for CBD stones, 315–316, 315f
intraoperative and postoperative choledochal cysts, 328
hemorrhage, 222–226 for hilar cholangiocarcinoma, 336–337,
definition of, 222, 223t, 224 339f
diagnosis and management of, 224 intraductal papillary mucinous
preoperative and intraoperative neoplasm, 55
considerations, 225–226 pancreas, 394–395, 395f, 396f
long-term benefits, 222 for pancreatic cysts, 55, 56f
outcomes/follow-up, 230–231 pancreatic ductal system, 393–394,
PHI and liver-related mortality 394f
definition of, 228–229 Magnetic resonance imaging (MRI)
diagnosis and management of, 229 for acute pancreatitis, 18, 18f
preoperative and intraoperative CRLM diagnosis, 179, 180
considerations, 229–230 focal nodular hyperplasia, 215
predictors of, 221, 222 gallstone disease, 288, 289f
Liver transplantation (LT) hemangiomas, 213
biliary leak mangement after, 364 hepatocellular adenoma, 216
complications of hepatocellular carcinoma, 389f–390f
bacterial and fungal infections, 380 liver, 388, 389f–390f, 390
biliary complications, 378, 379 pancreas, 394–395, 395f, 396f
HCC and PTLD, 380 pancreatic neuroendocrine neoplasms,
immunosuppressive drugs, 379, 380 67, 69f
primary nonfunction, 378 for PDAC, 42
thrombosis, 378 Magnetic resonance pancreatography
contraindications for, 373, 374 (MRP), for chronic pancreatitis, 29
donors expansion, 373 Main duct of Wirsung, 3, 4, 5f
factors, 373 Major biliary ductal injury, 355
with hepatocellular carcinoma, 373, Medical College of Wisconsin CT-based
374, 374t clinical staging, of pancreatic
hepatopulmonary syndrome, 374 cancer, 76, 77t
hepatorenal syndrome, 374 MEN 1. See Multiple neuroendocrine
indications for, 373, 374t neoplasia type 1 (MEN 1)
neuroendocrine carcinoma, 196, 197f Mesenteric drainage technique, 382, 383
002179469.INDD 442 7/14/2014 7:58:24 AM

Index 443
Metastatic colorectal cancer modified RECIST criteria (mRECIST),

ablation, 241 238
chemotherapy, 240 Monopolar electrocautery, 157, 158
drug-eluting beads, 241 Mucinous cystic neoplasms (MCNs), 53
stereotactic body radiation therapy, 241 Multidetector-row CT (MDCT),
surgical resection for, 240–241 388, 394
Yttrium-90, 241 for HCC, 170
Metastatic tumor, 44 for PDAC, 42
Metronidazole, 213 Multidisciplinary liver tumor group
Microwave ablation, 184, 239 HCC (See Hepatocellular carcinoma
Microwave scalpel, 163 (HCC))
Minimally invasive biliary surgery metastatic colorectal cancer
complications of ablation, 241
biliary fistula/stricture, 351, 352 chemotherapy, 240
hernias, 352 drug-eluting beads, 241
jejunojejunostomy leak, 352 stereotactic body radiation therapy,
laparoscopic vs. open biliary-enteric 241
reconstruction, 351t surgical resection for, 240–241
indications for, 346, 347, 347t Yttrium-90, 241
outcomes/follow-up, 351t, 352 Multiple cystic dilations, 326f, 327
postoperative management, 351–352, Multiple neuroendocrine neoplasia type
351t 1 (MEN 1), 62, 71
preoperative considerations
cholangiography and cross-sectional N
imaging, 346 NEC. See Neuroendocrine carcinoma
history taking and examination, (NEC)
345, 346 Necrosis, patterns of, 22, 23f
instrumentation, 346 Necrotic debris, endoscopy, 404
patient variables, 345 Necrotizing pancreatitis (NP)
Roux-en-Y hepaticojejunostomy endoscopic necrosectomy, 23
biliary-enteric anastomosis, intervention in, 21
350–351, 350f antibiotic treatment, 22
mobilization, 348–350, 349f goals of, 22
patient position, 347 patient selection, 22
port site placement for, 347, 348f percutaneous drainage, 23
side-to-side jejunojejunostomy, Neoadjuvant therapy, for PDAC, 45–46
347, 348 Neoplasms, pancreatic neuroendocrine.
Minimally invasive PD (MIPD), for See Pancreatic neuroendocrine
PDAC, 48 neoplasms
Minimally invasive surgery (MIS), for Neuroendocrine carcinoma (NEC),
pancreas 201–202
contraindications for, 89 ablative therapy, 198
indications for, 89 chemotherapy, 199–201
laparoscopic distal pancreatectomy, computed tomography, 191f, 192f, 193f
89–91, 90f diagnostic evaluation of, 188
laparoscopic fine needle aspiration/core biopsy,
pancreaticoduodenectomy, 91–93 190
outcomes of, 94–97 grades of, 190–191
postoperative management of, 93 tumor site of origin, 189
Minnesota tubes, 135 hepatic metastases from, 187–188
Minor bile leaks, 355 hormonal therapy, 199
Mirizzi syndrome, 296, 297f liver transplantation, 196, 197f
Model for end-stage liver disease (MELD) palliative resection, 194–196
score, 421 potentially curative resection of,
for cirrhosis, 113, 138 193–194
liver disease severity, 130 radiation therapy, 199
liver transplantation, 373 radiolabeled somatostatin receptor
Modified Milwaukee classification, for scintigraphy, 191
Sphincter of Oddi dysfunction, transarterial therapies, 198–199
300, 301t treatment follow-up evaluation, 201
Modified pancreatic classification Neuroendocrine tumors, transarterial
system, for Sphincter of Oddi therapy of, 427
dysfunction, 300, 301t Nifedipine, 305, 306
0002179469.INDD 443 7/14/2014 7:58:24 AM

444 Index
Nodular regenerative hyperplasia (NRH) pancreatic parenchyma, 151, 152f

diagnosis of, 218 target lesion, evaluation of, 153–154,
presentation of, 217–218 153f, 154f, 155f
treatment of, 218 vascular relationships, 153
Nonalcoholic fatty liver disease/ lymphatic drainage of, 9, 10f
nonalcoholic steatohepatitis MRI with MRCP, 394–395, 395f, 396f
(NASH), for HCC, 169, 233, 236 neck of, 2–3
Nonanatomic wedge resections. See PET scan, 395–396
Laparoscopic liver surgery physiology of, 10–13
Noncholestatic liver disease, 373, 374t ultrasound, 394
Nonselective shunts, 136–137 uncinate process, 2
NP. See Necrotizing pancreatitis (NP) venous blood supply, 7–9, 8f
Nutrition, for SAP management, 20–21 Pancreas transplant alone (PTA), 381
Pancreas transplantation
complications of
O
abscess/fistula, 383, 384
Octreotide, 70, 199
postoperative hemorrhage, 383
for acute variceal bleeding, 135
small bowel obstruction, 384, 385
for pancreatic neuroendocrine
thrombosis, 383
neoplasms management, 72
contraindications for, 381
Oncogenes, for PDAC, 40–41
immunosuppression, 383
Open biliary-enteric reconstruction vs.
indications for, 380–381
laparoscopic reconstruction, 351t
outcomes, 385–386
Open pancreatic debridement, for SAP
surgical technique, 381–383
management, 24–25
bladder drainage, 381, 382
Organ space infections
endocrine graft drainage, 382
definition of, 226
enteric exocrine drainage, 381
diagnosis and management of,
mesenteric drainage, 382, 383
226–227, 228f
tension-free enteric anastomosis,
preoperative and intraoperative
382, 382f
considerations, 227–228
Pancreatectomy
Orthotopic liver transplant (OLT)
central, 87
downstaging, 174
distal
HCC patient management, on waiting
description, 83–84
list, 174
hybrid approach, 86
Milan criteria, 173–174
laparoscopic approach, 86
Oven effect, ablative technique, 239
lateral to medial approach, 84–85
medial to lateral approach, 85
P pancreatic fistula incidence, 100
Packed red blood cells (PRBC), 224 splenic preservation, 85–86
PAIR (puncture, aspiration, injection, total, 86–87
reaspiration) technique, 210 Pancreatic cystic lesions
Palliative resection, in neuroendocrine benign, 51
carcinoma, 194–196 diagnosis of
Pancreas carcinoembryonic antigen level, 57
anatomy of, 2–3 clinical history in, 54
arterial blood supply of, 5–7, 6f cyst fluid analysis, 57–58
body and tail of, 3 cytopathology, 57
computed tomography, 394 DNA and biochemical markers, 58t
embryology, 3–5 physical examination in, 54
endocrine radiologic, 54–57
cells of, 11, 12t serologic, 54
function of, 11 symptoms of, 54
exocrine pancreas, functional unit IPMN (See Intraductal papillary
of, 11 mucinous neoplasm (IPMN))
fluid collections management of
pseudocyst, 402, 403 complications, 60
transmural drainage, 404 intraoperative considerations, 59–60
transpapillary drainage, 403, 403f operative indications, 58–59
head region, 2 postoperative management, 60
innervation of, 9–10 preoperative considerations, 59
intraoperative ultrasound, 151–154 outcomes and follow-up, 60–61
pancreatic duct, 152 premalignant/malignant cysts, 52–54
002179469.INDD 444 7/14/2014 7:58:24 AM

Index 445
Pancreatic duct anatomy, 3–5, 5f goals of, 100

Pancreatic ductal adenocarcinoma nutritional support, 103
(PDAC) percutaneous drain management,
classical vs. pylorus-preserving PD, 47 102f
diagnosis of, 41–43 somatostatin, 101
CA 19-9, 41–42 tubogram, 103f
clinical presentation, 41 Pancreatic injuries
imaging modalities, 42–43 complications, 414
laboratory analysis, 41–42 operative care, 413
serum markers, 42 postoperative management, 414
tissue markers, 42 preoperative considerations, 413
DNA mismatch repair genes, 41 technical tips, 413
etiology of, 39–41 Pancreatic intraepithelial neoplasia
gemcitabine, 49 (PanIN), 39
management of, 43–46 Pancreatic necrosis, 19f
assessing surgical resectability, 43–44 Pancreatic neuroendocrine neoplasms
diagnostic laparoscopy, 44–45 AJCC Classification System, 74t
neoadjuvant therapy, 45–46 diagnosis of
preoperative biliary drainage, 45 clinical presentation, 62, 63t–64t,
molecular genetics, 40–41 65–67, 66f
oncogenes, 40–41 imaging and localizing studies, 67,
outcomes and follow-up, 49–50 68f, 69f, 70
pathology of, 39 etiology of, 62
risk factors for gastrinomas, 65
acquired/environmental factors, 40 glucagonomas, 65
demographic factors, 39–40 incidence of, 62
host factors, 40 insulinomas, 65
surgical resection, 39 management of
adjuvant therapy, 48–49 localized disease, 70–71
classical vs. pylorus-preserving MEN1, gastrinoma arising, 71
PD, 47 metastatic disease, 71–72
distal pancreatectomy, 46–47 outcomes, 73, 73t, 74t, 75t
extended lymphadenectomy, 47 systemic therapy, 72–73
margin status, 48 nonfunctioning, 67
minimally invasive PD, 48 somatostatinomas, 67
pancreaticoduodenectomy, 46 VIPomas, 66
postoperative complications, 48 WHO classification, 75t
surgical palliation, 47 Pancreatic surgery
vascular resection/reconstruction, complications
47–48 biliary fistula, 107
tumor suppressor genes, 40 delayed gastric emptying, 103–105
variants of, 39 exocrine and endocrine insufficiency,
vascular resection/reconstruction, 107, 109
47–48 pancreatic fistula, 99–103
Pancreatic ductal system postoperative hemorrhage, 105–107,
computed tomography, 392–393 106f, 108f
endoscopic retrograde morbidity rate, 99
cholangiopancreatography, 394 technical aspects
magnetic resonance central pancreatectomy, 87
cholangiopancreatography, distal pancreatectomy, 83–86
393–394, 394f pancreaticoduodenectomy, 76–83
percutaneous transhepatic preoperative considerations, 76
cholangiography, 394 total pancreatectomy, 86–87
ultrasound, 392 Pancreatic tail mucinous cystic
Pancreatic fistula neoplasm, laparoscopic IOUS of,
definition of, 99 154, 155f
incidence of, 100 Pancreaticobiliary maljunction (PBM), 3
parameters for, 99, 100t Pancreaticoduodenectomy
risk factors for, 99–100 biliary anastomosis, 82
treatment of closure procedure, 82
drainage, 101 description, 76
endoscopic retrograde gastrojejunostomy, end-to-side, 82
cholangiopancreatography, 101 pancreatic anastomosis, 81–82
0002179469.INDD 445 7/14/2014 7:58:25 AM

446 Index
Pancreaticoduodenectomy (Continued) Pneumoperitoneum, laparoscopic

PDAC resection, 46 resection for HCC, 246
pylorus preservation, 82 Polysomy, 335
surgical resection steps, 76–80 Portal hypertension
infrapancreatic SMV isolation, 76, 77 ascites, 132
Kocher maneuver, 78 definition of, 126
ligament of Treitz, 78 esophageal varices, 134
portal dissection, 78, 79f laparoscopic resection for HCC, 246
sharp dissection, SMA margin, liver transplantation, 137–138
80, 81f preoperative considerations of, 138
terminal branches ligation, 78 spontaneous bacterial peritonitis, 134
traction sutures, 78–79 venous collaterals, 127, 128f
uncinate process removal, 79, 80f Portal vein embolization (PVE), 181–182,
venous resection, 83 181f
Pancreatitis, ERCP role in postoperative hepatic insufficiency, 230
acute, 402 preoperative (See Preoperative portal
chronic, 401–402 vein embolization)
Pancreatobiliary drainage, EUS for, 405 Portal veins and liver resections, 260,
Pancreatoblastomas, 39 262, 262f
Pancreatoduodenectomy (PD), for Portosystemic shunts
chronic pancreatitis, 31 nonselective shunts, 136–137
Parenchymal transection, for energy partial shunts, 137
devices. See Energy devices, for selective shunt, 137
parenchymal transection Positron emission tomography (PET)
PathFinder TG test, 57–58 CRLM diagnosis, 179
PDAC. See Pancreatic ductal of liver, 390
adenocarcinoma (PDAC) of pancreas, 395–396
Pedicle clamping, 168 for pancreatic cysts, 56
Percutaneous biliary intervention, in for PDAC, 43
biliary obstruction Posthepatectomy biloma, 228
immediate and delayed complications Posthepatectomy complications. See also
of, 419 Liver surgery, complications of
indications and contraindications to, case studies, 221, 222t
415, 416t definitions of, 223t
outcomes/follow-up of, 419 Posthepatectomy hemorrhage (PHH),
postprocedural management of, 418 definition of, 222, 223t, 224
surgical procedure, 415–418 Posthepatectomy liver failure. See
Percutaneous drainage Postoperative hepatic insufficiency
iatrogenic bile duct injuries, 419 (PHI)
of postoperative abdominal fluid Postoperative hemorrhage
collection, 427 complications of pancreas
for pyogenic liver abscess, 211, 212 transplantation, 383
for SAP management, 23, 24 liver surgery, complications of
Percutaneous transhepatic definition of, 222, 223t, 224
cholangiography (PTC), 415, 417f diagnosis and management of, 224
bile leak, 227 preoperative and intraoperative
biliary ductal system, 394 considerations, 225–226
choledochal cysts, 328 postpancreatectomy, 105–107, 106f, 108f
pancreatic ductal system, 394 Postoperative hepatic insufficiency (PHI)
Periampullary malignancy, ERCP role in definition of, 228–229
distal bile duct stricture, 399, 399f diagnosis and management of, 229
tissue sampling, 400 phases of, 221
Peri-pancreatic inflammatory preoperative and intraoperative
collections, IOUS, 154, 155f considerations, 229–230
Peripherally inserted central catheter Postoperative intra-abdominal fluid
(PICC), 104, 105 drainage, 427–429
Peutz-Jeghers syndrome, 40 Postoperative pancreatic fistula (POPF),
Piggyback technique, for LT, 375, 377f, 48
378 grading system for, 31, 35t
PlasmaKinetic tissue management Postpancreatectomy
system, 161–162 definition, 105
Plate-sheath system of liver, 263–265, incidence of, 105
264f, 265f treatment, 105–107
002179469.INDD 446 7/14/2014 7:58:25 AM

Index 447
Potentially curative resection, 193–194 endoscopic retrograde

Preoperative biliary drainage (PBD), for cholangiopancreatography, 407
obstructive jaundice, 45 side-to-side jejunojejunostomy, 348
Preoperative portal vein embolization, Roux-en-Y choledochojejunostomy, 323,
422–423, 423f 364, 404
Primary CBD stones, 314 Roux-en-Y hepaticojejunostomy
Primary nonfunction (PNF), 378 for BDI repair, 367, 368f
Primary sclerosing cholangitis (PSC), 333 minimally invasive biliary surgery
Prophylactic cholecystectomy, biliary-enteric anastomosis,
indications for, 283t, 284 350–351, 350f
Pseudoaneurysms (PSAs) mobilization, 348–350, 349f
computed tomography angiography, outcomes/follow-up, 351t, 352
429f, 430 patient position, 347
magnetic resonance angiography, port site placement for, 347, 348f
429f, 430 side-to-side jejunojejunostomy,
management of, 430 347, 348
prevalence of, 429 Roux-en-Y jejunal limb, 31, 32f
Pseudocyst, 402–404 Roux-en-Y jejunal loop, 34f, 36f
Pylorus-preserving
pancreatoduodenectomy (PPPD), S
31, 104 SAP. See Severe acute pancreatitis (SAP)
Pyogenic liver abscesses (PLA), 211–212, Sclerotherapy, for acute variceal
212f bleeding, 135
Secondary common bile duct
R stones, 314
Radiation therapy, for neuroendocrine Secondary malignancies, in liver. See
carcinoma, 199 Metastatic colorectal cancer
Radical antegrade modular Sectionectomies
pancreatosplenectomy (RAMPS) anatomic resections of liver, 258f
procedure, 47 left lateral sectionectomy, 247, 250f
Radioembolization, 237–238, 426–427 Segmentectomies, 259f
Radiofrequency ablation (RFA) Selective arterial stimulation, 70
colorectal liver metastases, 184 Selective internal radiation therapy
complications of, 424 (SIRT), 198–199
contraindications of, 424 Selective shunts, 137
goal of, 423 Self-expandable metallic stents (SEMS),
hepatocellular carcinoma, 216 400, 401
indications of, 424 Sengstaken-Blakemore tubes, 135
postoperative care, 425, 425f Serous cystadenomas (SCA), 51, 52f
preoperative workup, 424 Serous cystic neoplasms (SCNs), 51, 52f
Radiofrequency-assisted parenchymal Serum C-reactive protein (CRP), for AP
transection (RF-PT), 162–163, 166 severity, 15–16
Radiolabeled somatostatin receptor Severe acute pancreatitis (SAP). See also
scintigraphy, 191 Acute pancreatitis (AP)
Ranson’s criteria, for AP severity, 15, 16t management of
Resectable tumor, 43 antibiotics for, 20
Retroperitoneal approach, for SAP bleeding, 21
management, 24 combined endoscopic and
Revised Atlanta criteria, for AP severity, percutaneous approach, 23
16, 17t endoscopic necrosectomy, 23
RFA. See Radiofrequency ablation (RFA) ERCP indications, 21
Right hepatectomy, 247–251, 251f fine needle aspiration, 21
Ringer solution, 20 fluid resuscitation of, 20
Robotic liver surgery, 248–249 ischemia, of colon/gallbladder, 21
Robotic-assisted laparoscopic pancreatic laparoscopic transabdominal
resection, 94, 96 debridement, 24
advantages and disadvantages of, 96t necrotizing pancreatitis (See
Rome III classification system, for Necrotizing pancreatitis (NP))
Sphincter of Oddi dysfunction, nutritional support, 20–21
302, 303t open pancreatic debridement, 24–25
Roux limbs percutaneous drainage, 23
biliary–enteric anastomosis, 350, 351 retroperitoneal approach, 24
biliary–enteric reconstruction, 347 transgastric debridement, 24
0002179469.INDD 447 7/14/2014 7:58:25 AM

448 Index
Severe acute pancreatitis (SAP) Stereotactic body radiation therapy

(Continued) (SBRT)
venous thromboembolism, 21 hepatocellular carcinoma, 238
volume resuscitation/ICU metastatic colorectal cancer, 241
monitoring of, 20 Stewart-Way classification, of common
natural history of, 14, 19–20 bile duct injury, 357t
Side-to-side jejunojejunostomy, 347, 348 Strasberg classification, of biliary leak,
Simple cyst excision procedure, 354, 356f
choledochal cysts, 329 Sunitinib, for pancreatic neuroendocrine
Simple pancreatic cysts, 51 neoplasms, 72, 73
Small bowel obstruction (SBO), 384–385 Surgical shunt creation, 136–137
SOD. See Sphincter of Oddi dysfunction Symptomatic gallstones, 283
(SOD) Synchronous carcinoid metastasis, 194,
Solid and cystic pseudopapillary 194f–195f
neoplasms (SPNs), 53–54 Systemic inflammatory response
Solid epithelial neoplasms, of exocrine syndrome, 226
pancreas, 39 Systemic therapy, for pancreatic
Solid lesions, benign. See Benign solid neuroendocrine neoplasms, 72–73
liver lesions
Somatostatin, 199 T
for pancreatic fistula treatment, 101 Technetium HIDA scan
99
for pancreatic neuroendocrine bile leaks, 360, 361

neoplasms management, 72–73 choledochal cysts, 328
somatostatin receptor scintigraphy, TheraSphere™ glass microspheres, 237
68, 70 TIPS. See Transjugular intrahepatic
Somatostatinomas, 67 portosystemic shunt (TIPS)
Sonography. See also Ultrasound (US) TissueLink, 161
for liver evaluation, 387 Total laparoscopic
for pancreatic evaluation, 394 pancreaticoduodenectomy (TLPD),
Sorafenib, for HCC, 176, 239 94, 95t
Space of Disse, 126 Total pancreatectomy, 86–87
Sphincter of Oddi dysfunction (SOD) Total pancreatectomy with islet
clinical presentation of, 302–303 autotransplantation (TP-IAT)
complications and prevention of, 310 for chronic pancreatitis, 36–37
description of, 300 Transabdominal ultrasonography (TAUS)
diagnostic workup and treatment of, for acute pancreatitis, 16–17, 17f
310f for CBD stones, 315
differential diagnosis of, 303–305 for PDAC, 42
epidemiology of, 300–301 Transarterial chemoembolization
modified Milwaukee classification for, (TACE), 198, 199
300, 301t Transarterial embolization (TAE),
modified pancreatic classification 198–199, 216
system for, 300, 301t Transarterial therapy, 426–427
pathogenesis and pathology of, 301–302 bland embolization, 426
Rome III classification system for, chemoembolization, 426
302, 303t of hepatic metastatic colorectal cancer,
SOM performance, 304–305 426–427
therapeutic approach for and hepatocellular carcinoma, 426
balloon dilation and stenting, and intrahepatic cholangiocarcinoma,
308–309 427
Botulinum toxin injection, 309 neuroendocrine carcinoma, 198–199
endoscopic sphincterotomy, of neuroendocrine tumors, 427
306–308, 307t postoperative care, 427
medical therapy for, 305–306 preprocedure workup, 427
post biliary sphincterotomy, failure radioembolization, 426
of, 309–310 Transduodenal sphincteroplasty, for CBD
surgical therapy for, 306 stones, 323
transduodenal biliary sphincteroplasty, Transgastric debridement, for SAP
with transampullary septoplasty, management, 24
306 Transgastric endoscopic debridement,
Standard bicaval anastomosis, for LT, of NP, 23
375, 377f, 378 Transjugular intrahepatic portosystemic
Steatotic HCA, 216 shunt (TIPS)
002179469.INDD 448 7/14/2014 7:58:25 AM

Index 449
for acute variceal bleeding, 136 Vasopressin administration, for acute

indications and contraindications to, variceal bleeding, 135
421, 421t Venous blood supply, pancreas,
in-shunt stenosis, 137 7–9, 8f
model for end-stage liver disease score, Venous resection, 83
421 Venous sampling, 70
postoperative care of, 422 Venous thromboembolism, 21
preoperative evaluation of, 422 Verner-Morrison syndrome of
schematic illustration of, 419, 420f VIPomas, 67
success rates of, 421 VIPomas, 66, 67
Transmural drainage, 404
Transpapillary drainage, 403, 403f
W
Trauma, biliary leak mangement after,
Washout phenomenon, 388
364–365
Water-Jet dissector, 164
Trisomy 7, 335
WDHA syndrome. See Verner-Morrison
Tumor suppressor genes, 40
syndrome of VIPomas
Tumor-node-metastasis (TNM) schema,
Wedge resections. See Laparoscopic liver
73, 74t
surgery
Two-stage hepatectomy, 182, 182f
Whipple procedure. See
Pancreaticoduodenectomy
U Whipple triad, insulinomas
Ultrasound (US) diagnosis, 65
acute pancreatitis, 16–17, 17f Whittaker test, 418
biliary ductal system, 392 World Health Organization (WHO)
choledochal cysts, 327, 328 classification
cystic echinococcosis, 207, 208f differentiation of cysts, 207, 209t
gallbladder, 390–391, 391f of pancreatic neuroendocrine
gallbladder disease, 286–287, 287f neoplasms, 73, 75t
hemangiomas, 213
liver, 387–388
pancreas, 394 Y
pancreatic ductal system, 392 Yttrium-90 (Y90)
advantages of, 237
metastatic colorectal cancer, 241
V
microspheres, 240
Vancomycin, 275
Vardenafil (Levitra), 305
Vasoactive intestinal polypeptide (VIP), Z
66–67 Zollinger-Ellison syndrome, 65
0002179469.INDD 449 7/14/2014 7:58:25 AM

Handbook of Hepato-Pancreato-Biliary Surgery PDF

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

Handbook of Hepato-Pancreato-Biliary Surgery PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Handbook of Hepato-Pancreato-Biliary Surgery PDF

Uploaded by

Copyright:

Available Formats

Handbook of

0002164308.INDD 1 7/15/2014 12:39:34 PM

0002164308.INDD 3 7/15/2014 12:39:34 PM

Library of Congress Cataloging-in-Publication Data

002164308.INDD 4 7/15/2014 12:39:34 PM

0002164308.INDD 5 7/15/2014 12:39:34 PM

002164308.INDD 6 7/15/2014 12:39:34 PM

0002164308.INDD 7 7/15/2014 12:39:34 PM

Mashaal Dhir, MBBS

Jeffrey A. Drebin, MD, PhD

Evan L. Fogel, MSc, MD, FRCP(C)

002164308.INDD 8 7/15/2014 12:39:34 PM

Jonathan A. Fridell, MD, FACS

Wesley D. Leung, MD, MS

0002164308.INDD 9 7/15/2014 12:39:35 PM

Edward Lin, DO, MBA, FACS

Richard S. Mangus, MD, FACS

Robert C.G. Martin II, MD, PhD

002164308.INDD 10 7/15/2014 12:39:35 PM

Katherine A. Morgan, MD, FACS

Attila Nakeeb, MD, FACS

Timothy M. Pawlik, MD, MPH, PhD, FACS

Charles H.C. Pilgrim, MBBS (Hons), PhD

0002164308.INDD 11 7/15/2014 12:39:35 PM

Juan M. Sarmiento, MD, FACS

C. Max Schmidt, MD, PhD, MBA, FACS

Shimul A. Shah, MD, MHCM

002164308.INDD 12 7/15/2014 12:39:35 PM

Mark J. Truty, MD, MSc

Charles M. Vollmer Jr, MD

0002164308.INDD 13 7/15/2014 12:39:35 PM

002164308.INDD 14 7/15/2014 12:39:35 PM

put out fires that arise; best description is comprehensive brevity.

0002164308.INDD 15 7/15/2014 12:39:35 PM

Over the past decade, hepato-pancreato-biliary (HPB) surgery has

002164308.INDD 16 7/15/2014 12:39:35 PM

1 Pancreatic Anatomy and Physiology 2

5 Pancreatic Cysts Including Intraductal Pancreatic

6 Pancreatic Neuroendocrine Neoplasms 62

7 Technical Aspects of Pancreatic Surgery 76

8 Minimally Invasive Pancreatic Surgery 89

9 Complications of Pancreatic Surgery 99

0002164308.INDD 17 7/15/2014 12:39:35 PM

SECTION II: LIVER 110

10 Liver Anatomy, Physiology, and Preoperative

11 Cirrhosis and Portal Hypertension 126

12 Intraoperative Ultrasound of the Liver, Bile Ducts,

13 Energy Devices for Parenchymal Transection

14 Hepatocellular Carcinoma 169

15 Management of Patients with Colorectal Liver

16 Neuroendocrine and Noncolorectal Liver

17 Benign Liver Tumors 204

18 Complications of Liver Surgery Including

19 Multidisciplinary Approach to Liver Oncology 233

20 Laparoscopic Liver Surgery 243

002164308.INDD 18 7/15/2014 12:39:35 PM

SECTION III: BILIARY 253

SECTION II: LIVER 110

SECTION III: BILIARY 253

21 Liver and Biliary Anatomy 254

22 Infections in Biliary Surgery 271

23 Gallbladder Disease 279

24 Sphincter of Oddi Dysfunction 300

25 Common Bile Duct Stones and Exploration 314

26 Choledochal Cyst 325

27 Biliary Malignancies 332

28 Minimally Invasive Biliary Surgery 345

29 Bile Leak and Bile Duct Injury 354

SECTION IV: MULTIDISCIPLINARY HPB CARE 372

a nterior–posterior abdominal trauma. The gastroduodenal artery passes